IMPORTANT NOTICE

Recognizing that Alaska Gold Company (AGC), a subsidiary of NovaGold Resources, Inc. has legal and regulatory obligations in a number of global jurisdictions, AMEC E&C Services Inc (AMEC) consents to the filing of this report with any stock exchange and other regulatory authority and any publication by AGC, including electronic publication on AGC’s website accessible by the public, of this report.

This report was prepared as a National Instrument 43-101 Technical Report, in accordance with Form 43-101F1, for AGC by AMEC. The quality of information, conclusions, and estimates contained herein is consistent with the level of effort involved in AMEC’s services, based on: i) information available at the time of preparation, ii) data supplied by outside sources, and iii) the assumptions, conditions, and qualifications set forth in this report. This report is intended to be used by AGC, subject to the terms and conditions of its contract with AMEC. That contract permits AGC to file this report as a Technical Report with Canadian Securities Regulatory Authorities pursuant to provincial securities legislation. Except for the purposes legislated under provincial securities laws, any other use of this report by any third party, is at that party’s sole risk.

Harry M. Parker
AMEC Engineering & Construction Services, Inc.
780 Vista Blvd., Suite 100
Sparks, Nevada, 89434-6656
USA
Telephone: 775 353 3485 ext 132
Fax: 775 331 4153
Harry.parker@amec.com

CERTIFICATE OF AUTHOR

I, Harry M. Parker, do hereby certify that:

1.	I am Technical Director of AMEC Engineering and Construction Services, Inc. Sparks, Nevada
2.	I graduated with a BSc. in Geology, MSc. in Statistics and PhD in Geology from Stanford University in 1967, 1974 and 1975 respectively. In addition I have obtained an A.M. degree in Geology from Harvard University in 1969.
3.	I am a Fellow of the Australasian Institution of Mining and Metallurgy and The Society of Economic Geologists. I am a Member of the Institution of Mining, Metallurgy and Materials, the Australian Institute of Geoscientists, International Association for Mathematical Geology, Society for Mining, Metallurgy and Exploration, Canadian Institution of Mining, Metallurgy and Petroleum and the Geological Society of America
4.	I have worked as a geologist for a total of 39 years since my graduation from university.
5.	I have experience in resource estimation for vein and stockwork-hosted gold deposits, including Fort Knox, AK, True North AK, Juneau, AK, San Gregorio, Uruguay, Kupol, Russia, Achyem, Ghana, and Colomac NWT.
6.	I have read the definition of “qualified person” set out in National Instrument 43-101 (“NI- 43-101”) and certify that by reason of my education, affiliation with a professional association (as defined in NI 43-101) and past relevant work experience, I fulfill the requirements to be a “qualified person” for the Purposes of NI 43-101.
7.	I am responsible for the preparation of all sections of the “Technical Report of the Rock Creek Property, Nome, Alaska, USA” dated September 10th, 2006 (“the “Technical Report”) relating to the Rock Creek Property. I visited the property in June and September 2004 for 2 days. I have had prior involvement with the property that is the subject of the Technical Report. This involvement includes review of exploration data in 2003 and resource estimates prepared during January to March 2004.

AMEC
780 Vista Boulevard, Suite 100
Sparks, Nevada 89434-6656
USA
Ph: 1.775.331.2375		P:\P151088\9.0 Report\8.2 Final
Fax: 1.775.331.4153	www.amec.com	Report\Certificate of Authorhp_sep10.doc

8.	I am independent of the issuer applying all of the tests in Section 1.4 of National Instrument 43-101.
9.	I have read National Instrument 43-101 and Form 43-101F1, and the Technical Report has been prepared in compliance with that instrument and form.
10.	I consent to the filing of the Technical Report with any stock exchange and other regulatory authority and any publication by them for regulatory purposes, including electronic publication in the public company files on their websites accessible by the public, of the Technical Report.
11.	As of the date of this certificate, to the best of my knowledge, information and belief, the Technical Report contains all scientific and technical information that is required to be disclosed to make the Technical Report not misleading.
Dated this 10th of September, 2006.

Harry M. Parker

AMEC
780 Vista Boulevard, Suite 100
Sparks, Nevada 89434-6656
USA
Ph: 1.775.331.2375		P:\P151088\9.0 Report\8.2 Final
Fax: 1.775.331.4153	www.amec.com	Report\Certificate of Authorhp_sep10.doc

 

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

C O N T E N T S

1.0	SUMMARY		1-1
	1.1	Scope of Work	1-1
	1.2	Location	1-2
	1.3	Land Status	1-2
	1.4	Drill Data	1-2
	1.5	Geology and Mineralization	1-3
	1.6	Metal-at-Risk	1-4
	1.7	Mineral Resources	1-5
	1.8	Comment on Resources	1-6
	1.9	February 2005 Reconciliation	1-7
	1.10	Conclusions	1-7
	1.11	Recommendations	1-8
2.0	INTRODUCTION		2-9
	2.1	Terms of Reference	2-9
3.0	RELIANCE ON OTHER EXPERTS		3-1
4.0	PROPERTY DESCRIPTION AND LOCATION		4-1
	4.1	Location	4-1
	4.2	Mineral Tenure	4-1
	4.3	Permits, Agreements and Taxes	4-5
	4.4	Legal and Environmental Requirements	4-8
5.0	ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE, AND PHYSIOGRAPHY		5-1
	5.1	Accessibility	5-1
	5.2	Regional Centers and Infrastructure	5-1
	5.3	Climate	5-2
	5.4	Physiography	5-4
	5.5	Earthquake/Tsunami Risk	5-4
6.0	HISTORY		6-1
7.0	GEOLOGICAL SETTING		7-1
	7.1	Regional Geology	7-1
	7.2	Local and Property Geology	7-2
	7.3	Structural Geology	7-4
8.0	DEPOSIT TYPES		8-1
9.0	MINERALIZATION		9-1
	9.1	Mineral Zone Construction	9-2
	9.2	Alteration	9-5
10.0	EXPLORATION		10-1
11.0	DRILLING		11-1

Project No. 151088	TOC i
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	11.1	Drillhole Summary		11-1
	11.2	Review of 2004 Site Visit		11-2
		11.2.1	RC Drilling	11-2
		11.2.2	RC Logging	11-3
		11.2.3	Diamond Drilling	11-3
		11.2.4	Coring	11-3
		11.2.5	Core Logging	11-3
	11.3	Data Orientation		11-4
	11.4	Data Summaries by Sample Campaign		11-8
12.0	SAMPLING METHOD AND APPROACH			12-1
	12.1	Introduction		12-1
	12.2	RC Sampling		12-1
	12.3	RC Recovery and Data Selection		12-2
	12.4	Core Sampling		12-4
	12.5	Core Recovery and Data Selection		12-4
	12.6	Review of Twin Hole Data Selection		12-9
	12.7	Downhole Coefficient of Variation		12-14
	12.8	Trenching		12-17
	12.9	Sludge Sample Examination		12-18
13.0	SAMPLE PREPARATION, ANALYSES, AND SECURITY			13-1
	13.1	Laboratory Sample Procedures		13-1
	13.2	SGS Check Assay Program Results		13-2
	13.3	Review of Sampling Quality		13-4
14.0	DATA VERIFICATION			14-1
	14.1	Drillhole Survey Check		14-1
	14.2	Database Check		14-1
15.0	ADJACENT PROPERTIES			15-1
	15.1	Big Hurrah Deposit		15-1
16.0	MINERAL PROCESSING AND METALLURGICAL TESTING			16-1
	16.1	Ore Types		16-1
	16.2	Metallurgical Test Work		16-1
	16.3	Ore Processing		16-3
17.0	MINERAL RESOURCE AND MINERAL RESERVE ESTIMATES			17-1
	17.1	Gold Model Definition		17-1
	17.2	Drill-Hole Data Preparation		17-1
	17.3	Unadjusted Assay Statistics		17-3
	17.4	Adjusted Assay Statistics		17-3
	17.5	Adjusted Composite Statistics		17-4
	17.6	Contact Plots		17-5
	17.7	Declustering		17-5
	17.8	Quantile-Quantile (Q-Q) Plots (General)		17-7
	17.9	Twin Site Q-Q Plots		17-9
	17.10	Q-Q Plot Adjustments		17-10
	17.11	Comparison of RC and Core Data		17-10
	17.12	Adjustment of RC to Core Datum		17-11

Project No. 151088	TOC ii
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	17.13	Observations	17-12
	17.14	Variograms for Assays	17-13
	17.15	Variograms of 5 m Composites	17-14
	17.16	Development of Target Coefficients of Variation for Resource Model	17-15
	17.17	Gold Grade Interpolation Runs	17-15
	17.18	Metal-at-Risk	17-17
	17.19	Metal-at-Risk Adjustments	17-19
	17.20	Density	17-20
	17.21	Resource Classification	17-21
	17.22	Model Validation	17-22
	17.23	Resource Statements	17-28
	17.24	Reconciliation to Previous Models	17-30
18.0	OTHER RELEVANT DATA AND INFORMATION		18-1
19.0	INTERPRETATIONS AND CONCLUSIONS		19-1
20.0	ADDITIONAL REQUIREMENTS FOR TECHNICAL REPORTS ON DEVELOPMENT PROPERTIES AND PRODUCTION PROPERTIES		20-1
21.0	RECOMMENDATIONS		21-1
22.0	REFERENCES		22-1
23.0	DATE AND SIGNATURE PAGE		23-1

T A B L E S

Table 1-1:	Mineral Inventories (Model 4 is the Base Case Recommended for Planning)	1-6
Table 5-1:	Monthly Average Temperatures and Precipitation for Nome, Alaska (NovaGold website) Period of Record: 9/1/1949 to 9/30/2005	5-3
Table 6-1:	Rock Creek Drill Summary in Metres (Avalon, 2002)	6-2
Table 9-1:	Mineral Zones	9-2
Table 10-1:	Significant Drilling Results from Reconnaissance Exploration Prospects in the TNR – NovaGold Joint Venture Area (St. George, 2000)	10-5
Table 11-1:	Rock Creek Sampling Campaigns	11-1
Table 11-2:	Drillhole Sizes	11-4
Table 11-3:	Data Orientation	11-4
Table 11-4:	Data Inclination	11-5
Table 11-5:	Data Inclination by Type	11-7
Table 11-6:	Assays by Sample Campaign (rejected samples omitted)	11-8
Table 12-1:	Core Recovery Statistics	12-6
Table 12-2:	Subdivided Core Recovery Statistics	12-8
Table 12-3:	Twin Hole Data	12-11
Table 12-4:	Twin Hole Scores	12-12
Table 13-1:	Assay Quality Control Samples	13-2
Table 15-1:	Big Hurrah Mineral Resource	15-2

Project No. 151088	TOC iii
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 16-1:	Gold Recovery by Designated Metallurgical Process at a P80 of 212 microns (Norwest Corp, 2004)	16-3
Table 17-1:	Resource Model Limits	17-1
Table 17-2:	MineSight® Geologic Objects	17-2
Table 17-3:	Unadjusted Gold Assays by Mineral Zone	17-3
Table 17-4:	Adjusted Gold Assays by Mineral Zone	17-4
Table 17-5:	Adjusted Declustered Gold 5 m Composites by Mineral Zone	17-4
Table 17-6:	Declustered Statistics (Au g/t) for Assay Distributions	17-7
Table 17-7:	Variogram Models for Assays (AURJ1 g/t)	17-13
Table 17-8:	Derivation of Nugget Effects for Composite Variogram Models	17-14
Table 17-9:	Variogram Models for 5 m Composites	17-15
Table 17-10:	Grade Models	17-17
Table 17-11:	Metal-at-Risk by Mineral Zone (Indicated and Inferred)	17-18
Table 17-12:	Metal-at-Risk Removal	17-19
Table 17-13:	Density by Rock Type	17-20
Table 17-14:	Density by Depth from Surface	17-21
Table 17-15:	Indicated Resource Criteria	17-22
Table 17-16:	Kriged Validation Against Nearest Neighbour Models (Indicated Only)	17-23
Table 17-17:	Mineral Inventories (Model 4 is the Base Case Recommended for Planning)	17-29

F I G U R E S

Figure 2-1:	Metric Units of Measure and Abbreviations	2-10
Figure4-1:	Map Showing Bering Straits Lease Area, AGC Claim Area and Roads (AGC)	4-2
Figure 4-2:	Claim Map of the Rock Creek Project (AGC)	4-3
Figure 5-1:	Nome Location and Road Infrastructure	5-2
Figure 6-1:	Map Showing Drillhole Locations up to year 2000(Avalon, 2002)	6-2
Figure 7-1:	Regional Geology of Seward Peninsula, Alaska (Avalon, 2002)	7-2
Figure 7-2:	Local Geology of Rock Creek Deposit (Avalon, 2002)	7-3
Figure 8-1:	Tension Veining in Zone 1, Looking Northeast (AGC)	8-2
Figure 9-1:	Isometric View of Rock Creek Mineral Zones-Looking North (AGC)	9-4
Figure 10-1:	Exploration Targets in the Rock Creek Area (Avalon, 2002)	10-4
Figure 11-1:	Drillhole and Trench Location Map (5 m topographic contour interval)	11-6
Figure 12-1:	RC Recovery vs. Gold Grade	12-3
Figure 12-2:	RC Recovery vs. Hole Depth	12-3
Figure 12-3:	Core Recovery vs. Unadjusted Gold Grade	12-6
Figure 12-4:	Figure: Core Recovery Statistics	12-7
Figure 12-5:	Core Recovery vs. Down-hole Depth (Moving Average)	12-7
Figure 12-6:	Twin Hole Drill Sites	12-10
Figure 12-7:	Group A, CV vs Mean Gold	12-15
Figure 12-8:	Group B, CV vs Mean gold	12-15
Figure 12-9:	Group C, CV vs. Mean Gold	12-16
Figure 12-10:	CV vs. Mean Gold-Twin Holes	12-16

Project No. 151088	TOC iv
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 12-11:	CV vs. Mean- All Holes	12-17
Figure 15-1:	Big Hurrah Mineral Zone Perspective	15-2
Figure 16-1:	Process Flow Diagram	16-4
Figure 17-1:	Sample Q-Q Plot Before Adjustment	17-8
Figure 17-2:	Sample Q-Q Plot After Adjustment	17-8
Figure 17-3:	Twin Hole Q-Q Plot	17-9
Figure 17-4:	316 vs. 512 Twin Hole Q-Q Plot	17-10
Figure 17-5:	Example Profile – Gold Grade vs. Elevation	17-24
Figure 17-6:	Example Profile – Gold Grade vs. Northing	17-25
Figure 17-7:	Example Profile – Gold Grade vs. Easting	17-26

A P P E N D I C E S

A	Data Review
	A-1	Legal Opinion on Land Status by Guess & Rudd
	A-2	Certified Collar Coordinate Locations
	A-3	Rock Creek Sampling and Assay Methods
	A-4	Rock Creek Data Review Report, 28 June 2003
	A-5	Memo Reports – 2003 QA/QC, 2004 QA/QC
	A-6	Memos – Rock Creek Database Validation Updated, 13 February and 16 November 2004
	A-7	Memos – Chemex Sample Preparation Facility and Analytical Lab Inspections
	A-8	Memo – Estimation of Selection Bias: Rock Creek Metallic Screen Fire Assays for Gold above Selected Cutoff, 20 February 2004
	A-9	Downhole Variograms for Various Core Recovery Bins
	A-10	Twin Hole Data Grade Profiles
	A-11	Rejected RC Assay Intervals
	A-12	Memo Report – H. M. Parker 2004 Site Visit
	A-13	Downhole RC & DDH Composites > 1.0 g/t
B	Declustering
	B-1	Declustered Frequency Distributions
	B-2	Q-Q Plots
	B-3	Special Runs Made for Robert Prevost
C	Histograms & Statistics
	C-1	Assays Histograms & Probability Plots
	C-2	Composites Histograms & Probability Plots
	C-3	Contact Plots
D	Variography
	D-1	Assays Variograms

Project No. 151088	TOC v
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	D-2	Composites Variograms
E	Q-Q Plot Adjustments, Kriging and Classification Scripts
	E-1	Assay Adjustment Batchfiles
	E-2	Kriging Batch and Runfiles
	E-3	Mineral Classification Batch and Runfiles
F	Block Model Validation
	F-1	Block Model Histograms
	F-2	Nearest Neighbor Estimate Histograms
	F-3	Swath Plots
	F-4	Herco
	F-5	Block Model Cross Sections and Plan Maps
G	Resource Statements
H	Calculation of Estimation Variance for Rock Creek Bulk Sample
I	Rock Creek Reconciliation Memo

Project No. 151088	TOC vi
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

1.0	SUMMARY
1.1	Scope of Work
	Alaska Gold Company (AGC), a subsidiary of NovaGold Resources, Inc. (NovaGold) has asked AMEC E&C Services Inc (AMEC) to provide a mineral resource estimate and Qualified Person’s report for the Rock Creek gold deposit, located near Nome, Alaska, USA.
	The work entailed preparation of a mineral resource estimate in conformance with the CIM Mineral Resource and Mineral Reserve definitions referred to in National Instrument (NI) 43-101, Standards of Disclosure for Mineral Projects. It also involved the preparation of a Technical Report as defined in NI 43–101 and in compliance with Form 43–101F1 (the “Technical Reports”).
	AMEC conducted several site visits, the most recent being in September 2004.
	AMEC prepared a report in mid-2004 that contained the results of exploration conducted up to 2003 and resource models prepared in the spring of 2004. This report serves as an update and includes the results of exploration conducted during the summer of 2004 and associated resource models.
	AMEC and AGC personnel jointly, are responsible for the resource modeling. Stan Dodd and Robert Prevost of AGC are responsible for initially defining mineral zones, topography and faults. John Odden of AGC updated their interpretation to include the summer 2004 drilling. Jack Cote of NovaGold provided databases used to support resource modeling. For AMEC, Kevin Francis led the resource modeling effort, prior to his present employment with NovaGold, with the assistance of Mike Lechner and Dr. Harry Parker. Scott Long of AMEC analyzed the QA/QC results to determine their suitability in resource estimation.
	The resource models were prepared in Tucson, Arizona during the period October- December 2004. Mintec’s MineSight® software was used for most of the work; variography was performed using SAGE2001; some specialized AMEC software was also used.
	The work represents a disclosure of mineral resources for the Rock Creek deposit. Dr. Harry Parker, P. Geo., served as the Qualified Person for this report. Dr. Parker made a site visit on the 2nd of September 2004, and directed the mineral resource estimate and review of the geological data.

Project No. 151088	Page 1-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

1.2	Location
	The Rock Creek Project, located on the Seward Peninsula in the State of Alaska, is situated about 13 km north of the town of Nome, Alaska, 870 km northwest of Anchorage, and 160 km south of the Arctic Circle.
1.3	Land Status
	Guess and Rudd, an Alaskan law firm based in Anchorage retained by AGC, have issued a Limited Title Report dated February 7, 2006 and a letter dated August 11, 2006 concerning titles to the property and rights of Alaska Gold. The letter notes a mining sublease between AGC and Golden Glacier, Inc dated April 13, 2006 and a surface use agreement between Sitnasuak Native Corporation and AGC dated May 26, 2006. Guess and Rudd are of the opinion that “the execution of these agreements grants sufficient rights to Alaska Gold Company to enable it to extract gold and other minerals from the Lands”. These agreements satisfy two of the comments and requirements listed in the Limited Title Report. Guess and Rudd also stated “we reasonably expect Alaska Gold Company will satisfy, to the extent necessary in due course, all of the remaining comments and accompanying requirements set forth in our Title report dated February 7, 2006.” In a letter dated August 14, 2006, AGC confirmed with AMEC its intention to satisfy the comments and accompanying requirements.
	Based on this information, AMEC believes there is sufficient evidence for AGC to publicly disclose a resource at Rock Creek.
1.4	Drill Data
	Rock Creek has been explored by reverse circulation (RC) and core holes over a 16- year period. Currently, there are 25,175 m of core drilling and 12,935 m of RC drilling. There are also 1,490 m of trenching, located predominately in Zone 99 outside the main area of economic interest. In 2004, approximately 1,000 m of trenching occurred between the Walsh and Tension Vein Zones.
	A grade versus core recovery plot shows loss in gold (>10 g/t) where the core recovery is lower than 60%. Zones of lower core recovery are concentrated within 10 m of the surface and near the north-dipping Sophie Gulch Fault Zone. Because of possible low grade bias, sample intervals with core recovery less than 60% were excluded for the purposes of resource modeling. This resulted in 5.7% of the assay intervals for core samples being rejected.

Project No. 151088	Page 1-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	Core and RC data were compared by means of 132 twins, 110 of which were completed during 2004. The grade profiles for the twin holes seldom correlate well in detail. This relates to the narrow-veins containing gold mineralization. Cumulative grade thickness plots were created, and these were compared and qualitatively graded A through F, with A or B being the thickest. The coefficient of variation (CV) was found to decrease as the qualitative pair grades became poorer. A decrease in CV is often indicative of mixing of sample in the hole resulting from contamination from above. An equation was developed relating CV and grade for suspicious intervals. Decay and cyclicity studies were performed, but were not very useful, given a 10 ft (3 m) rod length, producing only two samples per rod, for RC holes drilled prior to 2004.
	RC data were studied on sections and compared to adjacent core holes. Unusually thick, low CV intervals were excluded as possibly contaminated. Approximately 2.6% of the RC metres sampled were excluded.
	The RC and core data were declustered and compared by drilling campaigns. In general, core data have between 20% and 40% lower grades than RC data, with the greatest differences occurring where the early RC campaigns are compared to core holes. Equations were developed to permit adjustment of RC distributions to core “datum.” These equations were developed using 2003 and 2004 Summer Campaign assay data, considered to provide the most reliable core samples. The equations were checked by subdividing the data by campaign, distance below topography, core recovery, and location inside or outside the Albion Shear Zone.
	Steep RC holes were drilled early on. It is difficult to tell whether the greater discrepancy between grades for steep RC and core holes is related to campaign or inclination. The adjustment equations are validated when RC data are compared to core samples with 100% recovery. Depth below topography does not appear to have a bearing on core and RC differences to a depth of 75 m. Below 75 m, the data available are limited. Within the Albion Shear Zone, RC and core assays compare reasonably well; therefore, no adjustment has been applied to RC assays in that zone.
	In a limited area, RC and core data were compared with trenches. Trench grades are on average 25% higher than those for core data. The trench grades average 11% lower than the RC grades.
1.5	Geology and Mineralization
	The Rock Creek deposit is underlain by bedrock consisting of several metamorphic rock sequences (the Nome Group), each of which has undergone at least two tectonic

Project No. 151088	Page 1-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	events. The protoliths are believed to be Cambrian to Devonian sediments consisting of shales, siltstones, sandstones, marls and limestones, deposited in a shallow continental shelf setting (Norwest, 2003). The bedrock is overlain by a variety of unconsolidated sediments. If frozen, these sediments are mostly thaw-unstable, especially on steeper slopes.
	The Nome group appears locally as a progression sequence consisting of four major units. The basal unit is complexly deformed pelitic schist, with a mixed unit consisting of mafic, pelitic, and calc schists and marble forming the next unit. The third unit is a mafic dominated schist, and the top unit consists of impure marble.
	Quartz muscovite schist (QMS) is the dominant rock in the Rock Creek Project area, but other types of schist are also present.
	Gold mineralization at the Rock Creek Deposit is developed in two styles: tension veins with quartz and varying amounts of carbonate, arsenopyrite and pyrite, and shear-hosted, within the Albion shear zone.
	Mineralization at Rock Creek occurs as relatively coarse, native gold occurring along fractures and veins, making the deposit amenable to recovery by gravity and flotation processes. The mineralized zone covers an area roughly 500 m wide and 1,500 m long, and as currently modeled, the pit will reach 100 m in depth. The individual mineralized zones within the deposit typically range from 10 to 50 m in width.
1.6	Metal-at-Risk
	Precious metal deposits have skewed grade distributions. Skewed grade distributions have the property that a small proportion of samples can represent a disproportionately large amount of contained metal. The limited number of these samples can introduce significant uncertainty into a resource estimate. It is common practice to cut the grades of very high-grade samples, restrict their projection distance, or to adjust resource models to mitigate downside risk.
	At Rock Creek 0.6% to 0.8% of the composites represent between 6% and 18% of the contained metal. On average, only three to six of these composites will be mined in a given year.
	To evaluate metal-at-risk, AMEC ran a Monte Carlo simulation that effectively re-drills the deposit 1,000 times and computes the high-grade metal associated with each realization. The 20th percentile for high-grade metal content is determined. The

Project No. 151088	Page 1-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	difference between the metal represented by the observed high-grade distribution and the metal content for the 20th percentile of the simulations is termed metal-at-risk. If this amount of metal is adjusted out of the resource model, four years out of five the mine can be expected to do better and one year out of five not as well.
	In the main zones of interest (1, 2, 10), between 6% and 10% of the metal was determined to be at risk. In the more sparsely drilled Zones 3 and 99, approximately 18% of the metal is determined to be at risk.
1.7	Mineral Resources
	Resource modeling was performed using ordinary kriging, with validation models being prepared using a nearest neighbour approach. The base-case model was constructed using composites adjusted to core datum (AUCO). This model was adjusted to give CVs for 10 × 10 × 5 m blocks that are similar to those expected for 10 × 10 × 5 m selective mining units (SMUs); these were chosen by AGC for an envisioned 7,000 t/d open-pit operation. The kriging plan involves an initial pass to assign grades in poorly drilled areas using a minimum of two composites per hole, and a maximum of eight, with no more than two per hole. In a second pass, eligible blocks are re-kriged with a minimum of three composites, and a maximum of eight, with no more than two composites being used from one drillhole, thus forcing the use of two drillholes where they are within the minimum search distance.
	For Zone 99, the grades of low- and high-grade populations were estimated. These were averaged using a kriged indicator (0.4 g/t) threshold to estimate the proportion of each population present in a block.
	Where required, metal-at-risk adjustment was achieved by capping the composite grades beyond a specified block-to-composite distance.
	The models were validated using swath plot comparisons of kriged and nearest neighbour model grades. Swaths were made by northing, easting and elevation.
	Resources were classified according to hole spacing in the vicinity of a block. In general, the area drilled on a 30 m grid is considered Indicated. The remaining blocks within 50 m of a drillhole (5 m) composite are considered Inferred.
	Table 1-1 shows mineral inventories for the resource models.

Project No. 151088	Page 1-5
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 1-1: Mineral Inventories (Model 4 is the Base Case Recommended for Planning)

Au	Resource	Indicated (Ind)			Inferred (Inf)			Contained Metal Ratio vs.
Cutoff	Model							Base Case Model (i.e.
(g/t)		Tonnes	Au	Au Ozs	Tonnes	Au	Au Ozs	Model 4)
		(000)	(g/t)	(000)	(000)	(g/t)	(000)	Indicated	Inferred
	Model 1	12,656	1.43	582	4,751	1.03	157	1.28	1.37
	Model 2	12,341	1.31	520	3,727	1.01	121	1.15	1.05
0.60	Model 3	11,369	1.39	508	4,608	1.03	153	1.12	1.33
	Model 4	11,040	1.28	454	3,583	1.00	115	1.00	1.00
	Model 5	14,052	1.45	655	4,938	1.03	164	1.44	1.43
	Model 6	13,783	1.34	594	3,918	1.00	126	1.31	1.10

Model 1 – Unadjusted Grades – no metal-at-risk removed
Model 2 – Unadjusted Grades, metal-at-risk removed
Model 3 – RC adjusted to core, no Metal-at-risk removed
Model 4 – RC adjusted to core, metal-at-risk removed (Base Case)
Model 5 – Core increased 25 %, no RC adjustment, no metal-at-risk removed
Model 6 – Core increased 25 %, no RC adjustment, metal-at-risk removed

At the request of NovaGold, Mike Lechner of Resource Modeling Inc. (RMI) generated a Lerchs-Grossmann pit shell using a US$500/oz gold price and other parameters taken from a Norwest Updated Economic Review. Mineral Resources within this pit are declared using Model 4:

	•	9.595 million tonnes (Indicated) at a grade of 1.31 g/t containing 404,000 oz Au.
	•	1.432 million tonnes (Inferred) at a grade of 0.96 g/t containing 44,000 oz Au.
1.8	Comment on Resources
	The discrepancy between core and RC data is troubling in that it is difficult to confidently make an adjustment to either dataset. Overall, the core recovery is very good, and it is difficult to imagine how a great deal of gold could have been lost in the coring process (fines and flakes of gold not recovered in the core barrel).
	The RC data show some evidence of downhole contamination, but generally grade profiles are fairly sharp, and long tails of low-grade gold below high-grade zones are not evident.
	Despite 38,000 m of drilling, at a reasonable spacing for the deposit-type, it is impossible to confidently make a statement of the resources present. AMEC recommends using Model 4, which was calculated by removing the metal-at-risk and having adjusted the RC data to the core data as it is the most conservative of the models and is based on the current available information.

Project No. 151088	Page 1-6
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

1.9	February 2005 Reconciliation
	AGC requested AMEC complete a reconciliation of resources between the AGC 2003 PACK model and two subsequent AMEC resource models.
	The 2003 PACK model contains about 400,000 more contained gold ounces than the other AMEC models within the 2003 scoping study pit. Approximately half of the difference (200,000 ounces) is due to the fact that the 2003 scoping study pit was optimized using all estimated blocks from the 2003 PACK model, whereas subsequent drilling has shown mineralization is absent or of lower grade. The remaining difference can largely be attributed to an increase in the proportion of core drilling and adjustment of RC drillhole grades to give a similar distribution to the grades found in core holes (not done for the 2003 PACK model, and perhaps justifiably given the much smaller amount of core data available at the time). In addition, the November 2004 AMEC model was estimated using significantly more data than were available to construct the 2003 PACK model.
1.10	Conclusions
	The Rock Creek property constitutes a quartz vein hosted deposit which has been extremely difficult to sample and has necessitated an adjustment to the samples data, where core data has been adjusted to RC data.
	Gold mineralization at the Rock Creek Deposit is developed in two styles: tension veins with quartz and varying amounts of carbonate, arsenopyrite and pyrite, and shear-hosted, within the Albion shear zone.
	Mineralization at Rock Creek occurs as relatively coarse, native gold occurring along fractures and veins, making the deposit amenable to recovery by gravity and flotation processes. The mineralized zone covers an area roughly 500 m wide and 1,500 m long, and as currently modeled, the pit will reach 100 m in depth. The individual mineralized zones within the deposit typically range from 10 to 50 m in width.
	Mineral Resources are declared using Model 4 (base case):
	•	9.595 million tonnes (Indicated) at a grade of 1.31 g/t containing 404,000 oz Au.
	•	1.432 million tonnes (Inferred) at a grade of 0.96 g/t containing 44,000 oz Au.

Project No. 151088	Page 1-7
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

1.11	Recommendations
	•	Conduct sampling studies to confirm and determine the source of gold loss during sample preparation.
	•	Drilling density may be adequate in areas of Mineral Zone 99, adjacent to the Albion Shear Zone, to identify geologic controls and to incorporate them into geologic model.
	•	Review of the databases identified a small number of problems, mostly related to substitution of one assay for another. It would be useful to perform a 100% check of pre-2003 data and identify the appropriate assay certificates considered to furnish “final” values for each sample interval. The QA/QC data must also be compiled in an organized group of datasets. A summary document on all QA/QC programs should be prepared. This may affect final resource classification.
	•	AGC has reduced the significance of pre-1990 Placer Dome and Tenneco data (particularly RC data) with new RC and core drilling throughout the greater part of the proposed pit. AMEC commends this effort and recommends continuing the process on sections such as 405N where older RC drilling is still dominant.
	•	AMEC recommends that AGC follow the recommendations stated by Guess and Rudd, in a letter dated February 7, 2006 to satisfy the Limited Land Title for AGC.

Project No. 151088	Page 1-8
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

2.0	INTRODUCTION
	Alaska Gold Company (AGC), a subsidiary of NovaGold Resources, Inc. (NovaGold) asked AMEC E&C Services Inc (AMEC) to provide an independent mineral resource estimate and Qualified Person’s report for the Rock Creek gold deposit, near Nome, Alaska, USA.
	The work entailed estimation of mineral resources, in conformance with the CIM Mineral Resource and Mineral Reserve definitions referred to in National Instrument (NI) 43-101, Standards of Disclosure for Mineral Projects. It also involved the preparation of a Technical Report as defined in NI 43–101 and in compliance with Form 43-101F1 (the “Technical Reports”). Dr. Harry Parker, P. Geo. is the Qualified Person responsible for preparing this Technical Report.
	Information and data for the independent resource estimate were obtained from NovaGold personnel in Vancouver and from the project site near Nome, Alaska. Other information came from Norwest Corporation of Calgary, Alberta, who performed an updated economic review of the Rock Creek deposit in August, 2005.
	Pertinent geological data were reviewed in sufficient detail to prepare this document. Dr. Harry Parker, P. Geo., a Technical Director with AMEC, directed the mineral resource estimation work and review of the geological data. Dr. Parker most recently visited the site on the 2nd of September 2004. No significant field work has been done since then.
	Secondly, Steven Blower, P. Geo, of AMEC, visited the Rock Creek site on April 22 and 23, 2003 to review the diamond drilling procedures, geology, archived drill core, quality control procedures, and verification of the assay database.
	Finally, Arne Bakke CPG, performed a review of the ALS Chemex Preparatory Lab in Fairbanks on September 20, 2004. He is the former Chief Geologist for Fairbanks Gold, operator of the Fort Knox gold mine. Mr. Bakke’s review was commissioned by AGC, and AMEC has relied on his review in preparation of this document.

2.1	Terms of Reference
	All units of measure (see Figure 2-1) used in this report are in the metric system, except the section on climate (US origin), where all published data are quoted in degrees Fahrenheit for temperature and inches for precipitation measurements. Also,

Project No. 151088	Page 2-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

the contained metal quantities shown in the mineral resource summary tables are expressed in troy ounces.

Figure 2-1: Metric Units of Measure and Abbreviations

Above mean sea level	amsl
Annum (year)	a
Centimetre	cm
Cubic centimetre	cm3
Cubic metre	m3
Day	d
Days per week	d/wk
Degrees Celsius	°C
Dry metric ton	dmt
Gram	g
Grams per tonne	g/t
Greater than	>
Hectare (10,000 m2 )	ha
Hour	h (not hr)
Kilogram	kg
Kilograms per cubic metre	kg/m3
Kilograms per tonne	kg/t
Kilometre	km
Kilometres per hour	km/h
Kilometres squared	km2
Less than	<
Litre	L
Metre	m
Metres above sea level	masl
Metric ton (tonne)	t
Micrometre (micron)	µm
Milligram	mg
Millimetre	mm
Million	M
Million tonnes	Mt
Minute (plane angle)	'
Ounce	oz
Parts per billion	ppb
Parts per million	ppm
Percent	%
Pound(s)	lb
Second (plane angle)	"
Short tonnes	st
Specific gravity	SG
Square centimetre	cm2
Square kilometre	km2
Square metre	m2
Thousand tonnes	kt
Tonne (1,000 kg)	t
Tonnes per annum	t/a
Tonnes per cubic metre	t/m3
Tonnes per day	t/d

Project No. 151088	Page 2-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

3.0	RELIANCE ON OTHER EXPERTS
	AMEC has not reviewed the land tenure, nor independently verified the legal status or ownership of the properties or underlying option agreements. AMEC has relied on the opinion of Guess and Rudd Law offices contained in the current Limited Title Report (Appendix A-1) regarding the legal status and ownership of the properties.
	The results and opinions expressed in this report are based on AMEC’s field observations and the geological and technical data listed in the Appendices. While AMEC has carefully reviewed all of the information provided by AGC, and their consultants, and believes the information to be reliable, AMEC has not conducted an independent in-depth investigation to verify its accuracy and completeness.
	AMEC has relied on Michael Lechner for generation of a Lerchs-Grossmann pit shell within which Mineral Resources have been declared. AMEC has assumed Mr. Lechner is a qualified professional.
	The results and opinions expressed in this report are conditional upon the afore mentioned technical and legal information being current, accurate, and complete as of the date of this report, and the understanding that no information has been withheld that would affect the conclusions made herein. AMEC reserves the right, but will not be obliged, to revise this report and conclusions if additional information becomes known to AMEC subsequent to the date of this report. AMEC does not assume responsibility for AGC’s actions in distributing this report.

Project No. 151088	Page 3-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

4.0 PROPERTY DESCRIPTION AND LOCATION

4.1	Location
	AGC’s Nome properties occur partly on patented mining claims owned 100% by the Alaska Gold Company, a wholly owned subsidiary of NovaGold and partly on land controlled by the Bering Straits Native Corporation. The deposits fall within Parcels I and II, centred on latitude 64.57 degrees north and longitude 165.39 degrees west.
	Rock Creek is located on the south coast of the Alaskan Seward Peninsula, facing Norton Sound in the Bering Sea, about 13 km north of the township of Nome. Nome is situated 870 km by air northwest of Anchorage, and 160 km south of the Arctic Circle (see Figure 4-1).
4.2	Mineral Tenure

The project is partly located on 5,700 hectares (14,000 acres) of patented private land that is 100% owned by AGC, and partly on land owned by the Bering Straits Native Corporation. AGC holds an exploration and mining lease on approximately 8,100 hectares (20,000 acres) of Bering Straits Native Corporation lands, as well as a surface use agreement with Sitnasuak Native Corporation, the local Nome village corporation (see Figure 4-2).

In August, 2006 Guess and Rudd, lawyers of Anchorage, Alaska, provided AGC with legal opinion on the Rock Creek land status.

They completed limited research on the ownership of Parcels I and II, which contain the Rock Creek Project.

Parcel I

The claims within Parcel I (red outline on Figure 4-1) are held in title by Alaska Gold Company, a wholly-owned subsidiary of NovaGold. Parcel I is on patented land owned by AGC and has unrestricted use of the surface.

Specific claims are identified in Table 4-1

Project No. 151088	Page 4-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure4-1: Map Showing Bering Straits Lease Area, AGC Claim Area and Roads (AGC)

Project No. 151088	Page 4-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 4-2: Claim Map of the Rock Creek Project (AGC)

Project No. 151088	Page 4-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 4-1: Claims within Parcel I Situated in U.S. Mineral Survey 721 (AGC)

Claim Name	Patent Number
Francisco	316745
No. 1 Above Francisco	316745
No. 2 Above	316745
No. 3 Above	316745
No. 1 Sophie Gulch	316745
No. 4 Above	316745
Fractional Claim No. 4 ½ on Rock Creek	316745
No. 5 Above	316745
No. 6 Above Right Hand Branch	316745
Rock Creek Bench No. 4 Above	955746

Parcel II

The lands within U.S. Mineral Survey 721 known as Bench Claim No. 4 ½ Above Placer and additional lands as conveyed by Alaska Gold Company in 1983 are held in title by the Sitnasuak, and Bering Straits Native Corporations (green outline on Figure 4-1).

For Parcel 2 AGC has an exclusive surface use agreement with the Sitnausak Native Corporation for an initial term of 7 years and so long thereafter as mining, processing, or marketing operations are carried out in good faith. The Sitnausak Native Corporation will be paid US$70,000 annually during the term of the agreement, and US$900 per acre disturbed, adjusted annually for inflation. AGC will also pay any and all real property taxes and assessments related to Sitnausak’s interest levied by any municipality or governmental entity.

The conveyed lands are described as follows:

•	Contained within KRM T. 10 S., R. 34 W.,
•	Sec.14:All, excluding the lands within Mineral Surveys 332, 721, and 1835
•	Sec.23:All, excluding the lands within Mineral Surveys 403, 447, 715, 721, and 2271
On June 8, 2005, Guess and Rudd, an Anchorage-based law firm retained by AGC, stated their opinion on Limited Land Title for AGC in respect to the Rock Creek Gold Deposit. AMEC recommended that AGC follow Guess and Rudd’s recommendations to clarify title. A second letter, dated February 7, 2006 has updated their opinion of the same matter. Finally, a letter dated August 11, 2006 states that two agreements have been negotiated since the second letter: 1) a mining sublease between AGC and

Project No. 151088	Page 4-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	Golden Glacier, Inc dated April 13, 2006; and 2) a surface use agreement between Sitnasuak Native Corporation and AGC dated May 26, 2006. In its August 11, 2006 letter, Guess and Rudd are of the opinon that “the execution of these agreements grants sufficient rights to Alaska Gold Company to enable it to extract gold and other minerals from the Lands”. These agreements satisfy two of the comments and requirements listed in the Limited Title Report dated February 7, 2006. Guess and Rudd also stated “we reasonably expect Alaska Gold Company will satisfy to the extent necessary in due course all of the remaining Comments and accompanying Requirements set forth in our Title report dated February 7, 2006, or if Alaska Gold fails to satisfy any such comment or accompanying requirement, such failure is unlikely (due to almost certain termination or expiration of any noted prior agreement) to affect adversely the exercise by Alaska Gold Company of its rights in and to the Lands (whether acquired under or pursuant to the two agreements described above or under or pursuant to other instruments and agreements”. In a letter dated August 14, 2006, AGC confirmed to AMEC its intention to satisfy the remaining comments and accompanying requirements.
	Based on this information, AMEC believes there is sufficient evidence for AGC to publicly disclose a resource at Rock Creek.
	The letters and opinions of Guess and Rudd dated June 8, 2005, February 7, 2006, and August 11, 2006; and a letter from AGC to AMEC dated August 14, 2006 are provided in Appendix A-1.
4.3	Permits, Agreements and Taxes
	The following is a list of necessary permits required prior to construction, which is being coordinated by Bristol Environmental & Engineering Services Corp (Norwest, 2005).
	•	Wetlands Permit
	•	Air Quality Construction Permit
	•	Stream Crossings Permit
	•	Land Application of Pit Water Permit
	•	Class V Injection Well Permit
	•	Solid Waste Permit
	•	EPA Storm Water Discharge Permit
	•	Coastal Zone Management Permit
	•	Telecommunications Permits

Project No. 151088	Page 4-5
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

• Fire Marshall Permit

Additionally, various operating plans must be submitted to the proper regulating authorities and include:

• Plan of Operation

• Spill Prevention and Control Contingency Plan

• Reclamation Plan

• Hazardous Waste Plan

• Monitoring Plan

• Road Maintenance Agreements

• Transportation Plan

AGC has stated to AMEC that the following are permits on-hand:

• US Army Corps of Engineers 404 Wetlands Permit

• Alaska Department of Environmental Conservation (ADEC) 410 Certification (state certification of the federal 404 permit)

• ADEC Waste Management/Injection Well Permit

• Alaska Department of Natural Resources (DNR) Title 41 Fish Habitat Permit

• DNR Temporary Water Use Permit (6 sources, 6 separate permits)

• EPA Stormwater Discharges Associated with Construction Activity under a General NPDES Permit

• DNR Coastal Zone Management Certification

• Stream Reclassification Petition

Also, AGC has stated the outstanding permits:

• ADEC Air Quality Construction Permit

• EPA Title V Injection Well Permit

• EPA Stormwater Discharges Associated with Operations Activity under a General NPDES Permit

The following information regarding permits and agreements is taken from NovaGold’s website (NovaGold, 2006).

Project No. 151088	Page 4-6
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

NovaGold owns 313 mineral surveys made up of one or more patented claims in the Nome area through its wholly-owned subsidiary, Alaska Gold Company. These mineral surveys are fee simple and have no annual requirements. A significant proportion of the mineral resources are located on lands owned by Alaska Gold Company (NovaGold 2006).

Pursuant to an exploration and option agreement dated March 13, 2002, between Golden Glacier, Inc., and the Company, the Company acquired the rights to explore and develop the lode deposits on an additional 15,000 acres of mineral claims held by Golden Glacier Inc. pursuant to four mining leases from Bering Straits Native Corporation (“BSNC”) to Golden Glacier, Inc. Pursuant to the exploration and option agreement, Golden Glacier Inc. granted the Company a five year option to acquire a mining sublease. In order to maintain the option in effect, the Company agreed to make annual payments to Golden Glacier Inc. ranging from US$15,000 to US$25,000 and to complete annual work commitments ranging from US$50,000 to US$150,000. If the Company exercises its option (which it is entitled to do at any time provided the agreement is in good standing), the Company will be granted a mining sublease for 30 years or so long thereafter as there is mineral production from the claims. In order to maintain the sublease in good standing the Company must carry out minimum work requirements of US$250,000, adjusted for inflation. Golden Glacier Inc. is entitled to a 2.5% net smelter return royalty and a 5% net proceeds royalty from production from BSNC lands. The Company is required to pay advance minimum royalties of US$100,000 during each year of the sublease.

NovaGold is also a party to an exploration surface use agreement with Sitnasuak Native Corporation (Sitnasuak) and is negotiating a mining surface use agreement with Sitnasuak, which has resulted in a US$70,000 annual payment not offset by an advance royalty plus $900 per acre of disturbance.

The BSNC has a US$126,000 advance royalty offset against a 2.5% net smelter return (NSR) and 5% net proceeds. Dore bars refined on site are assumed to incur a charge of $2.00 per oz to cover transportation, insurance, refinery deductions etc, to point of sale. It is assumed that the net profits definition is the same as for the State of Alaska’s Mining License Tax.

Both of the final two agreements mentioned above carry requirement for sensitivity to subsistence and require reclamation to a safe and stable configuration.

The actual production from BSNC lands has been determined on an annual basis for this evaluation. It is estimated from the geological model that 6% of the annual gold production

Project No. 151088	Page 4-7
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	is from BSNC land. Since the remainder of production is from patented mining claims owned by AGC, it is assumed that no production is from state land and so the Alaska Production Royalty on production does not apply.
	All mining income in Alaska is subject to the Mining License Tax. This is a graduated tax based on the level of net income, reaching 7% above US$100,000. As a new mine, Rock Creek should qualify for the three and a half year new mine exemption from the Mining License Tax.
	Alaska income tax will also be reduced by an exploration incentive tax credit. It is assumed that exploration conducted by the company in 2003 qualifies for this tax credit.
	According to NovaGold, all AGC payments, royalties and taxes for the Rock Creek Property are in good standing as of the date of this publication.
4.4	Legal and Environmental Requirements
	AGC is required by the State to provide acid-base accounting (ABA) records for their tailing pond to confirm that it will not be acid generating. Also required are site specific neutralizing potential:acidic potential (NP:AP) ratios for all lithology types at Rock Creek. Ratios above 2:1 are considered net neutralizing and those below 2:1 are considered potentially acid generating. Finally, a site water balance must be submitted to the State.
	There are no pre-existing liabilities for AGC, however the area has been placer mined extensively for the past 100 years. Much of the area shows visible disturbance and there are tailings piles located throughout the region from historical mining.

Project No. 151088	Page 4-8
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

5.0 ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE, AND PHYSIOGRAPHY

5.1	Accessibility
	Rock Creek is located on the south coast of the Alaskan Seward Peninsula, facing Norton Sound in the Bering Sea, about 13 km north of the township of Nome. Nome is situated 870 km by air northwest of Anchorage, and about160 km south of the Arctic Circle. The Rock Creek Project is currently road accessible via the Glacier Creek Road and the State maintained Teller-Nome Highway, an all-weather paved and gravel road. The State of Alaska is constructing the Glacier Creek Road By-Pass, which will simplify the road access distance to site.
5.2	Regional Centers and Infrastructure
	The city of Nome has a year-around population of approximately 4,000 and serves as the logistical and administrative center for Western Alaska. Nome has daily commercial jet service and large container barge traffic service from June through October. While there is no road to Nome from elsewhere in Alaska, there are about 500 km of state-maintained roads around the city (see Figure 5-1).
	The city of Nome has provided electricity to past mining operations and has offered that service for future operations if necessary. Current generating capacity is based on three diesel generators. The current local power consumption in Nome is in the range of 4MW to 6MW. No camp facilities are required at the Rock Creek Project due to its close proximity to Nome, which is well serviced with accommodations.
	AGC has stated that they have the surface rights for the mining operation, availability of water, mining personnel and access to power generation that is adequate to supply future plans.
	In addition, NovaGold’s operating sand-and-gravel and land businesses give the company an important operational presence in the community.
	Nome has a modern airport with a 6,000 ft (1,800 m) asphalt runway and is serviced by Alaska Airlines and several private air taxis. The Nome Airport provides intrastate, national and international access to the city. Both daily scheduled jet service and charter air services are available.

Project No. 151088	Page 5-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

The Port of Nome is the only harbour for boat moorage and service in the region. Sea lanes are generally open from June through to October; Norton Sound freezes over in late fall, with thawing in mid-spring.

The Norton Sound Regional Hospital serves Nome and fifteen other area villages. It is well-equipped, with facilities that include a laboratory, emergency and intensive care, long-term care, out-patient and labour wards.

Figure 5-1: Nome Location and Road Infrastructure

5.3	Climate
	The area is characterized by cool summers and cold winters with relatively low annual precipitation, averaging less than 16 inches per year, mostly as rain in the summer.

Project No. 151088	Page 5-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Precipitation is light, the average rainfall being about 20 inches per year; snowfall averages 30–48 inches per year. Nome is located on the Bering Sea on the southern shore of the Seward Peninsula, which has a relatively dry climate, easily accessible terrain and some of the best infrastructure in the State of Alaska, with roads providing year-around access to the property. Discontinuous permafrost occurs throughout the Seward Peninsula, and during the winter and early spring, sea ice can cover much of the Bering Sea. Winter temperatures in Nome range from minus 5 to minus 15 F; summer temperatures fluctuate from 45 to 55 F. Average temperatures and precipitation are presented in Table 5-1.

Table 5-1: Monthly Average Temperatures and Precipitation for Nome, Alaska (NovaGold website)
                    Period of Record: 9/1/1949 to 9/30/2005

	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Annual
Average Max.
Temperature	13.9	13.5	17.6	26.8	42.4	53.6	57.3	55.7	48.5	34.2	23.2	14.1	33.4
(Fahrenheit)
Average Min.
Temperature	-1.6	-2.6	0.2	11.7	29.8	39.6	45.2	44.2	36.3	22.6	10.7	-0.8	19.6
(Fahrenheit)
Average Total
Precipitation	0.88	0.72	0.61	0.70	0.73	1.06	2.16	3.33	2.48	1.47	1.11	0.88	16.12
(inches)
Average Total
Snow Fall	9.5	8.3	7.1	6.6	2.1	0.1	0.0	0.0	0.4	4.5	11.2	10.1	59.9
(inches)
Average Snow
Depth (inches)	14	15	16	13	2	0	0	0	0	1	4	8	6

The moderating influence of the open water of Norton Sound on the Nome climate is effective from early June to about the middle of November. Storms moving through this area during these months result in extended periods of cloudiness and rain. There is a nearly continuous cloud cover during July and August. During the summer months the daily temperature range is very slight. The freezing of Norton Sound in November causes a rather abrupt change from a maritime to a continental climate. The majority of low pressure systems during this period take a path south of Nome, resulting in strong easterly winds, accompanied by frequent blizzards, with the winds later becoming northerly and reaching Nome across the colder frozen areas of northern Alaska. Temperatures generally remain well below freezing from the middle of November to the latter part of April, with January usually the coldest month of the year. Temperatures usually begin to rise near the end of February and continue to rise until they reach a maximum in July.

Precipitation reaches its maximum during the late summer months and drops to a minimum in April and May. Snow begins to fall in September, but usually does not

Project No. 151088	Page 5-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	accumulate on the ground until the first part of November. The snow cover decreases rapidly in April and May, and normally disappears by the middle of June. Snow depths in Nome have exceeded 70 inches.
	Wind speeds are usually less than 15 knots, except during some of the intense storms experienced by the region during August. Severe windstorms do occur, with winds over 70 mph recorded several times. Strong winds during the winter months, when there is snow cover, produce blowing snow conditions that severely hinder transportation in the area.
	In terms of the metric system, the average annual air temperature at the Rock Creek deposit is about -3°C, with summer temperatures ranging from +8°C to +15°C, and winter averages of -15°C.
	AGC has stated to AMEC that the mine at Rock Creek intends to operate all year long.
5.4	Physiography
	Low marshy flats occur along the coast at Nome. These give way to a series of foothills, with heights of 150 to 365 m (500 to 1200 ft), in an arc to the northwest and northeast. The terrain increases in ruggedness and height farther north, with the Kigluaik Mountains reaching a height of 1,500 m (5,000 ft). The ground along the coastal flats is swampy during the summer months, but is permanently frozen below a depth of less than a metre (2 to 3 ft). Elevations in the project area range from 100 m (300 ft) to over 300 m (1,000 ft) along the highest ridgelines. Vegetation in the Nome area consists mostly of grass and numerous small flowering plants.
5.5	Earthquake/Tsunami Risk
	The earthquake/tsunami risk assessment is taken from the City of Nome’s disaster plan document.
	Numerous faults are mapped onshore near Nome; most trend north to northeast. However, seismic planning studies place the Nome area in a comparatively low-risk category, with a ten percent probability of earthquakes measuring 3.0-4.5 on the Richter scale in a fifty-year period. There is no recorded damage from earthquakes in Nome.
	The ADES designates Nome as having a low probability of occurrence of a tsunami.

Project No. 151088	Page 5-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

6.0	HISTORY
	The early history of the development of the Rock Creek area is sourced from Hawley and Hudson (1996).
	Major alluvial placer gold deposits in the Nome mining district were discovered in September and October 1898 by John Brynteson, Jafet Lindeberg, and Erik O. Lindblom. The men were supported by an active Scandinavian community based at Golovin, who mainly came to the region for the Swedish Covenant Church, and by several Eskimos, including Gabriel Adams and Constantine Uparazuck. A mining district was formed in October, 1898, by Brynteson, Lindeberg, Lindblom, A.N. Kittlesen, Johan Tornensis, a Saami, and Gabe Price, who represented Charles D. Lane, an experienced mining man.
	Many of the best alluvial placers were located in 1898. In 1899, gold was discovered on Present Beach at Nome. A.H. Brooks of the U.S. Geological Survey then predicted that buried beach placer deposits would be found, a prediction that proved accurate when Second Beach was discovered in 1902. The Third Beach was discovered in 1904. The placers were first exploited mainly by hydraulic methods, and major ditches were constructed to support these operations. Some shallow thawed deposits were mined successfully by small dredges. In the 1920s, after invention of the coldwater thawing process, large dredges were brought into the country by Wendell P. Hammond, whose interests were later consolidated by the U.S. Smelting, Refining, and Mining Company.
	The Nome mining district is the second most important placer district in Alaska. From 1898 to 1993, more than 4,800,000 ounces of gold (150 metric tonnes) were produced, essentially all by placer methods and mostly from complex alluvial deposits or buried beach deposits. The district is also estimated to have produced more than 550,000 ounces (17.54 metric tonnes) of silver. Very small amounts of stibnite and scheelite were also produced.
	Lode exploration began shortly after discovery of the placers, but that was not significant until the 1980s when geologist R.V. Bailey reopened old workings in Rock Creek and discovered sheeted veins in upper Snow Gulch. His work was followed up by Placer Dome Inc., Tenneco, Newmont Mining Company, Kennecott Exploration Company and Alaska Gold Company. In 1999, the assets of the successors to the U.S. Smelting, Refining, and Mining Company were acquired by NovaGold Resources.

Project No. 151088	Page 6-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

The first modern exploration for lode gold deposits in the Rock Creek district began in 1986 when Aspen Gold consolidated the district land holdings and began exploration directed toward bulk mineable gold deposits. What is now the Rock Creek deposit was discovered shortly thereafter as a result of dozer trenching. Following Aspen Gold’s initial discovery in 1986, a succession of major mining companies including Placer Dome, Tenneco, Newmont and Kennecott optioned and explored the property from 1988 through 1994. Approximately 315 holes were drilled on the Rock Creek property, with 94 holes located in the actual deposit. This drilling includes 7,590 m of core drilling and 8,137 m of rotary drilling (Table 6-1) and drillhole locations up to year 2000 (Figure 6-1).

Table 6-1: Rock Creek Drill Summary in Metres (Avalon, 2002)

Company	Core	Rotary
NovaGold Kennecott Newmont Placer Dome	0 3,653     321 3,616	2,886 2,024        0 6,113
Total	7,590	11,023

Figure 6-1: Map Showing Drillhole Locations up to year 2000(Avalon, 2002)

Project No. 151088	Page 6-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

NovaGold Resources Inc. acquired the property by purchasing AGC from parent Mueller Industries in May 1999. In July 1999, NovaGold entered into an option agreement with Viceroy Resources and later that year completed six reverse circulation drillholes (2,886 m) to supply samples for metallurgical test work and to test the reliability of previous drilling. All holes intersected longer and higher-grade intervals of mineralization than indicated by previous drilling (St. George, 2000). Inadequate compressor volume and pressure in previous drill programs were thought to have caused under-reporting of gold grades. NovaGold completed 30 drillholes at Rock Creek in 2000 to increase the drill density in the area of the defined gold resource. No field work was carried out on the property in 2001.

On February 18, 2002 NovaGold Resources entered into a letter agreement on Rock Creek with Vancouver-based TNR Resources. TNR can earn a 49.9% interest in a joint venture by investing $10 million by the end of 2004 ($1M through December 2002, an additional $3M by December 2003 and an additional $6M by December 2004). After the earn-in, NovaGold and TNR could maintain or dilute their percentage interest according to a straight-line formula. Upon regulatory approval, TNR will issue 500,000 of its common shares to NovaGold. In order to reduce share dilution, TNR may elect to use debt financing rather than equity financing to cover capital expenditures or contract out mining operations as part of its $10M commitment to fulfill its financial obligation. AGC has stated to AMEC on April 21, 2006, that TNR Resources no longer has a stake in the Rock Creek property.

NovaGold reported that its precursor companies had undertaken soil sampling, geophysics and geologic mapping and various drilling campaigns from 1986 to 1999. A total of 33,000 m of core and rotary drilling was carried out by these companies on the property (of which 18,600 m was on the Rock Creek deposit), as well as metallurgical test work (NovaGold, 2006).

Project No. 151088	Page 6-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

7.0 GEOLOGICAL SETTING

7.1	Regional Geology
	Information on the regional geology of the area has been taken from NovaGold (2006), Ford and Snee (1996), Bundtzen et al, 1994), and Hawley and Hudson (1996).
	The Seward Peninsula is composed of a complex series of metamorphic rock packages which have been affected by large-scale accretionary, rotational and transcurrent events (Figure 7-1). The rocks of the Seward Peninsula consist of continental shelf, platform and margin assemblages ranging from limestone and dolomite through pelitic and psammitic protoliths to intermediate and basaltic volcanic rocks. These rocks form part of the Arctic composite terrane of Plafker and Berg (1994) and were originally deposited along the northern margin of ancestral North America. Subsequent blueschist followed by amphibolite facies metamorphism modified these rock packages. Cretaceous and Tertiary compressional and extensional deformation, mid-Cretaceous plutonic activity and large-scale Tertiary counter-clockwise rotation have further modified rocks of the Seward Peninsula.
	The rotational segment of the tectonic history took place in two stages. The first, caused by rifting and opening of the Canada Basin, took place from 130 to 100 million years ago (Ma) and corresponds to counterclockwise rotation of Arctic Alaska. Widespread blueschist facies metamorphism accompanied this major compressional tectonic event (Goldfarb, 1997; Till and Dumoulin, 1994). Near the end of this time period, a greenschist facies event extending from 120-90 Ma overprinted the Jurassic blueschist facies rocks. Greenschist facies metamorphism was accompanied by high- strain deformation resulting in a prominent low angle penetrative foliation, northwest- southeast mineral stretching lineations, and recumbent isoclinal folding of the earlier fabric. Plutonic activity began approximately mid-way through the greenschist event (Nokleberg et al, 1998). This plutonic activity was generated by northward subduction of the Farallon and Kula Plates along Alaska’s southern and southeastern margins (Flanigan et al, 2000).
	Convergence of North America and Eurasia from about 66 to 50 Ma caused what is now the Seward Peninsula to be offset to the south along a regional-scale apparent left-lateral fault (Plafker and Berg, 1994). It is uncertain how this event affected previously emplaced gold systems on the Seward Peninsula.
	Information on the regional geology of the area is sourced from NovaGold (2006), Ford and Snee (1996), Bundtzen et al, 1994), and Hawley and Hudson (1996).

Project No. 151088	Page 7-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 7-1: Regional Geology of Seward Peninsula, Alaska (Avalon, 2002)

7.2	Local and Property Geology
	The area is underlain by metamorphic rocks belonging to the Nome Group, which is thought to constitute a coherent lithostratigraphic succession of four major units (see Figure 7-2):
	•	a basal, complexly deformed pelitic schist,
	•	a "mixed unit" of mafic, pelitic, and calc schists and marble,
	•	a mafic dominated schist package, and
	•	an impure marble package.

Project No. 151088	Page 7-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 7-2: Local Geology of Rock Creek Deposit (Avalon, 2002)

Project No. 151088	Page 7-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	The protoliths of these rocks are thought to be Cambrian to Devonian sediments consisting of shales, siltstones, sandstones, marls and limestones, deposited on a shallow water continental platform. The mafic volcanics are probably younger mafic sills. Nome Group rocks have undergone at least two periods of metamorphism and accompanying deformation. The earliest metamorphism was a blueschist facies event, which is considered to be mid-Jurassic in age, having a minimum Argon/Argon age of about 120 million years. This event was synkinematic with a high strain deformation, which resulted in widely, distributed penetrative fabric and mesoscopic intrafolial isoclinal folds. Following the blueschist metamorphism came a greenschist facies overprint, which was also accompanied by high-strain deformation. This resulted in a prominent low angle penetrative foliation, northwest-southeast mineral stretching lineations, and recumbent isoclinal folding of the earlier fabric. The event is interpreted to have occurred during a period of north–south crustal extension from 120 to 90 Ma. Most lode gold deposits of the Seward Peninsula have age dates that fall within the metamorphic event (120 to 90 Ma) and are interpreted to have been formed during that event.
7.3	Structural Geology
	This summary of the regional structural setting of the Rock Creek project is summarized from NovaGold (2006).
	The rocks show several episodes of strong deformation. Outcrops, where original bedding can be observed, often exhibit strong folding within an apparently sub- horizontal bed. Many of the low-angle contacts are probably imbricated thrusts related to the Brooks Range Orogeny, which occurred between late Jurassic and Tertiary times. Foliation developed during the greenschist event is sub-horizontal. Late-stage antimony and weak gold mineralization is associated with the northeast high-angle faults. One of the most prominent is the northeast striking Anvil Fault. The Anvil Fault has a significant geophysical signature and probable strike-slip movement. Glacier Creek and upper Rock Creek occupy parallel faults.
	The Brynteson high-angle fault may also represent a major conduit for metamorphic fluids including hydrocarbons. Abundant carbon is seen in chips from bedrock samples obtained over the Brynteson fault zone by rotary drilling. The fault may have also mobilized gold; most gold occurrences east of the Snake River are located within a few kilometres of the Bryntenson Fault. Carbon flooding and gold mineralization are also associated with the Penny Fault, west of the Snake River. High angle northeast- striking tension fractures extend across the district. These tension fractures are related to northeast-striking faults, east–west compression of the Seward Peninsula, north–

Project No. 151088	Page 7-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

south extension, or a combination of these forces. The fractures occasionally host veins. These zones of parallel veining typically have a vertical extent of no more than 100 m (300 ft) and a lateral extent of 300 m or more (900 ft). Veins in tension fractures have been seen in the hanging wall of low-angle structures at Rock Creek and Lindblom. The vein density is from one to three veins per metre.

The Albion shear, which hosts the Albion veins, appears to offset all other structures. There is a significant offset of lithologies across the 30 m wide Shear Zone. The Albion veins exhibit banding and re-healed breccias and have not been intensely fractured like the tension veins. These characteristics indicate the shear was active during a different, perhaps shallower environment than that under which the tension veins were formed.

Low angle faults are common in the Nome District. The faults often contain remobilized carbon and could be pre- or synmetamorphic thrusts. There is a lack of compressive deformation, but this deformation could be overprinted with lithostatic compression and foliation. Thrust faults could be remobilized by detachment. The low-angle faults do show evidence of remobilization. A low-angle fault at Rock Creek has both a carbon/calcite component and a gouge/clay component.

Project No. 151088	Page 7-5
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

8.0	DEPOSIT TYPES
	The Rock Creek project is classified as hosting low-sulphide gold veins, analogous to deposit model 36a of Cox and Singer (1986). These deposits can be mined by open- pit or underground mining methods.
	Despite production of nearly 5 million ounces of alluvial gold from the Nome District since 1898, there has been no significant lode production from the district. Evidence presented by Goldfarb (1997) and St. George (2000) suggests gold mineralization at Rock Creek was derived from metamorphogenic processes, as indicated by fluid geochemistry. Mapping by Bundtzen et al (1994) classifies much of the country rock as metaturbidite, a common host for metamorphogenic gold deposits in the Juneau and Valdez Creek districts in southeastern and central Alaska (Goldfarb, 1997).
	According to AGC, Rock Creek does not fit into any specific model type and is best described as a sheeted vein system in combination with a deposit scale mineralized shear zone. Gold mineralization is believed to be contemporaneous with late Cretaceous crustal extension.
	Tension veins are northeast-trending, northwest-dipping veins that host the majority of gold at Rock Creek. Individual veins are usually less than 10 cm wide and average 3.8 cm wide. Vein density typically varies between 1.5 to 3 veins per metre in sheeted vein zones. Vein density increases proximal to the northeast trending Albion Shear Zone and is common in tension fractures in the hangingwall of low-angle structures. Sheeted tension vein zones appear more continuous along strike (up to 300 m) than down-dip (<100 m). The tension vein zones are strongest in the south-central part of the property where the entire width of veining is up to several hundred metres. This zone appears to be confined in a lateral extent by two north northeast-trending faults and vertically by the Sophie Gulch and Lower Sophie Gulch faults. Figure 8-1 below shows tension veining at Rock Creek.
	The shear zone, known locally as the Albion, is up to 30 m wide and trends at a 050o azimuth with a moderate to steep northwest dip. This type of geologic structure is well known as a deposit type for lode gold mineralization and is more thoroughly discussed in the next section on mineralization.

Project No. 151088	Page 8-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 8-1: Tension Veining in Zone 1, Looking Northeast (AGC)

Project No. 151088	Page 8-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

9.0	MINERALIZATION
	Three styles of mineralization are recognized on the project:
	•	Replacement bodies: albite, quartz, arsenopyrite, dolomite, and minor galena.
	•	Tension veins: quartz veins with albite, arsenopyrite, and minor lead sulfosalts.
	•	Shear hosted veins: quartz veins with pyrite and lead sulfosalts.
	These veins are contained within schists and generally strike 050 and dip steeply (70 degrees) to the southeast. They have a strike length of at least 500 m and die out at a depth of about 200 m.
	Each mineralization type is described in more detail below (NovaGold, 2006).
	Replacement Bodies
	There is probably more than one episode of replacement mineralization. The first occurred in a ductile environment and is characterized by large crystals of albite, quartz, arsenopyrite, and dolomite. These bodies occur in low angle structures and small fold noses. They can appear to be sill like bodies. Tension fracture veins cut these bodies at several locations, although occasionally the replacement bodies appear to become thicker when approaching a vein. The replacement bodies occur by themselves but are also fairly common in areas of vein mineralization. Metallurgical studies indicate the gold in replacement bodies occurs as fine free gold in association with sulphides.
	Tension Veins
	The tension veins are usually less than 10 cm (4 in) wide and average about 3.8 cm (1.5 in). The veins appear to have more strike length than vertical extent. The tension vein zones can be hundreds of metres wide.
	The veins contain quartz with varying amounts of carbonate, arsenopyrite and pyrite. Trace amounts of base metal sulfides, stibnite, free gold and sulfosalt minerals are distributed erratically throughout the tension veins. Typical tension veins at Rock Creek average 0.6% arsenopyrite and 0.6% pyrite (combined syngentic and hydrothermal) by volume as estimated from visual inspection of 2004 core. Native gold commonly occurs on fractures in tension veins. The majority of visible gold seen in drill core from 2004 occurs within vein quartz or associated with arsenopyrite.

Project No. 151088	Page 9-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	Shear-Hosted Veins
	The Albion Shear Zone is a large (up to 30 m wide), well-defined structural zone that trends 050o azimuth with a moderate to steep northwest dip. It is characterized by extensive shearing (mylonitization), but also includes local zones of gouge and brecciation. The Albion Shear Zone style of quartz veining is distinctly different from the tension veining. These veins are often wider (a few centimetres to greater than 3 m) than the tension veins. They are commonly sheared and broken, can be banded and locally include quartz-cemented breccias indicating shearing was active during a different, perhaps shallower environment than that under which the tension veins formed. The quartz is less fractured than quartz in tension veins and arsenopyrite, base metal sulfides and stibnite total less than 0.5% of the zone by volume. Free gold is locally present though it is less common than in tension veins. Fine-grained pyrite and lead sulfosalts give the veins a dark or bluish colour. There is no discernable zonation of sulfide species along strike or downdip within the Albion Shear Zone.
9.1	Mineral Zone Construction
	Mineral zones were updated using new drillhole data that became available from the 2004 drilling program. AGC project geologist John Odden reviewed the existing interpretation and updated the sectional polygons where warranted. The deposit contains distinct zones with different grades and/or geological controls. To honour these distinct zones during grade estimation, the mineral zone envelopes were interpreted on section and then linked in MineSight®, creating three-dimensional wire- frame solids that were then used to code drillholes and to assign mineral zone codes to the block model. Mineral zones for the Rock Creek model (RKCK15.DAT) are described in Table 9-1.
	Table 9-1: Mineral Zones

Label	Code	Name/Nominal Grade Domain	Notes
MINZN	1	+0.5 g/t	Generally east of Albion Shear - Tension
			Veining
MINZN	2	+0.25 g/t	Weak mineralization both east and west of
			Albion Shear
MINZN	3	Walsh Zone	Provided by Stan Dodd sections, unchanged
			by AMEC
MINZN	10	Albion Shear	Adopted from prior solids interpolation,
			unchanged by AMEC
MINZN	99	Miscellaneous Mineralization	All other miscellaneous mineralization falls
			into MINZN 99

Project No. 151088	Page 9-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Mineral Zone 1 was initially defined by AGC geologists in an area located immediately east of the Albion Shear Zone that is characterized by sheeted tension veins. In general, this modeled zone was defined by material above a 0.5 g/t cutoff. As modeled, this zone does contain drillhole intercepts less than the 0.5 g/t gold cutoff, but in general, this zone shows good continuity between drill sections.

Mineral Zone 2 consists of three modeled solids that are located on either side of the Albion Shear Zone and is characterized by less continuous mineralization than Zone 1. This zone was initially modeled using a nominal 0.25 g/t gold cutoff grade, but like Zone 1, lower grade intervals were included.

Mineral Zone 3, located on the eastern side of the deposit, is identified as a series of three northeasterly trending sub-parallel zones also known as the Walsh Zone. Cross sectional interpretations were prepared by John Odden and linked by AMEC into wire-frames.

The Albion Shear Zone, or Mineral Zone 10, is a narrow, north-south trending high-angle zone of mineralization that has been delineated by a number of core and reverse circulation drillholes. AGC personnel reviewed all core drillhole logs and prepared a comprehensive list of intercepts that define this zone. John Odden then modified the previously constructed sectional polygons using the list of drillhole intercepts and newly acquired drillhole data. Those polygons were then linked together by AMEC creating a wire-frame solid. The Rock Creek Mineral Zones are shown (see Figure 9-1) in a perspective view with the drillholes for reference.

Project No. 151088	Page 9-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 9-1: Isometric View of Rock Creek Mineral Zones-Looking North (AGC)

The remainder of the deposit defaulted to Mineral Zone 99, which has no wire-frame boundaries and in general, is characterized by scattered intercepts of discontinuous mineralization. There are several exceptions, particularly west of the Albion Shear Zone around Section 540N, where some continuity of shear-style mineralization is recognized.

The sectional polygons for Mineral Zones 1 and 2 were locally modified by AMEC to overlap the Albion Shear Zone to prevent inadvertent gaps in the drillhole and model codes. These overlaps were created where Zones 1 and 2 were supposed to be in contact with the Albion Shear Zone. The drillholes and block model were sequentially coded with the wire-frames starting with Mineral Zone 3, then proceeding with Zones 2, 1, and finishing with Mineral Zone 10.

The sectional polygons and three-dimensional wire-frames that were used to code the drillholes and block model are stored in a folder within the MineSight® project directory called “Nov2004 Geo Interpretations”. The sectional polygons and final wire-frames are all stored under individual folders for each of the Mineral Zones.

Drillhole assay and composite intervals were backcoded (‘speared’) within MineSight® to assign codes for intervals that fell within at least 50% of each wire-frame solid. The

Project No. 151088	Page 9-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	MineSight® assay, composite, and block model items to receive those codes is called MINZN (Mineral Zone).
9.2	Alteration
	Extensive alteration, mainly albitization and sulfidization, accompanied early ductile lode mineralization (Hawley and Hudson, 1996). Arsenopyrite was the main introduced sulfarsenide. Later alteration during the brittle stage included sericitization, silicification, and ankeritization. The Albion Shear Zone has shown evidence of mylonitization along shear surfaces, along with localized brecciation and gouge occurrences.

Project No. 151088	Page 9-5
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

10.0	EXPLORATION
	Placer miners commonly found and reported gold bearing quartz veins throughout the Seward Peninsula but were unable to mine them profitably due to their discontinuous nature. Little lode exploration was conducted after a few failed attempts at underground mining during the early part of the century.
	In 1986, R.V. Bailey came to Nome with bulk mineable targets in mind and quickly consolidated the district land holdings. Rock Creek was discovered by trenching shortly thereafter. Rock Creek has been worked on by several different mining companies since 1986.
	During each of these work projects, soil sampling, geophysics, geologic mapping and various drilling campaigns were conducted. Work on the property was carried out subsequent to 1986, and prior to the NovaGold’s involvement by Placer Dome, Tenneco, Newmont, Kennecott and Alaska Gold. A total of 33,000 m of core and rotary drilling was carried out by these companies on the property (of which 18,600 m was on the Rock Creek deposit), as well as metallurgical test work.
	In June 1999, after NovaGold concluded the purchase of AGC, NovaGold began a review and compilation of all previous work in the Nome area. In September of that year, six reverse circulation drillholes (437 m) were completed for metallurgical test work, and these holes also tested the reliability and continuity of previous drilling. All six holes intersected significant widths of greater than 3 g/t gold mineralization.
	In 2000, 30 additional drillholes totalling 2,449 m were completed. Sampling was carried out on each 1.52 m (5 ft) interval, and strict sample protocols were employed to ensure adequate sample size and quality. The NovaGold’s program used experienced reverse circulation drillers employing a compressor with 900 CFM capacity, with the air-lift to ensure complete sample recovery. NovaGold stationed a geologist on the rig full-time to ensure strict sampling protocols were undertaken. The drilling program and sampling protocol were managed by NovaGold with oversight provided by Phillip St. George, then Vice-President, Exploration, for NovaGold. According to NovaGold, Phillip St. George is a Qualified Person as defined by National Instrument 43-101.
	In 2002, NovaGold completed 16 HQ diamond core drillholes totalling 1,182 m in a program funded by TNR Gold Corp. This drill program was designed to expand the extent of the known gold resource and to complete infill core drilling along the higher- grade Albion Zone of the deposit to support preparation of an independent preliminary economic assessment report.

Project No. 151088	Page 10-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

In 2003, a 30,000 ft (9,100 m) infill feasibility delineation drill program was completed, and a feasibility study was begun. Environmental baseline monitoring was also started in 2003. The 2003 drilling program confirmed that the coarse gold component of the mineralization at Rock Creek produced differing results for samples obtained from core and rotary drilling.

In 2004, NovaGold worked with AMEC E&C Services Inc. to design a comprehensive twinning and infill drilling program to define a final resource model, provide additional material to carry out further metallurgical test work to optimize the recovery of gold, and allow the NovaGold to develop the best approaches to grade control and mining methods.

In late 2005, a grade control study was initiated to better optimize planning for grade control. Results from that study will become available in the second quarter of 2006.

In addition to Rock Creek, a number of other prospects with anomalous gold and associated metals (arsenic-antimony) have been defined within the TNR – NovaGold joint venture area by soil sampling and drilling (Figure 10-1). These prospects cover a 10 km (6 mile) long area extending from the Goodluck target in the north to the Third Beach target in the south. Within this area a total of 87 drillholes have intersected significant gold mineralization. These 87 drillholes contain 123 discrete intercepts of +2 g/t over 1.5 m (5 ft) or greater (Table 10-1). Most intercepts have not been offset.

One of the most advanced of the targets outside of Rock Creek proper is the Saddle prospect, where limited drilling has extended the mineralized area over a 500 m (1,500 ft) distance along a northeast striking quartz-vein stockwork system similar to Rock Creek (St. George, 2000). Placer Dome estimated a resource of 544,320 tonnes at 5.76 g/t, and Tenneco estimated 1.32 million tonnes at 3.29 g/t. Mineralization is open along strike in both directions.

The Bonanza Hill prospect area has three zones of northeast trending mineralization all containing multiple drillholes with significant intercepts (Table 10-1). The northwestern portion of Bonanza Hill returned significant intercepts in an area measuring 1.2 km long and open to the northeast and southwest. The central zone at Bonanza Hill returned some of the most promising holes and trends into the Saddle zone, 1.47 km to the northeast. The southeast zone at Bonanza Hill is 630 metres long, with significant intercepts that are open to the southwest.

Several other prospects in the TNR – NovaGold joint venture area have returned promising drilling and/or soil sampling results (Table 10-1, Figure 10-1). At the

Project No. 151088	Page 10-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Goodluck prospect, one drillhole intersected 20.8 g/t over 3 m. This hole has not been offset. The Balto and Prospect Creek target areas are untested soil anomalies showing over 100 ppb gold. Lindblom Creek has been tested with only two holes which returned 3 m of 1.8 g/t and 6.1 m of 1.2 g/t. No other drilling has been done on this target. The Lindblom Pit area is a large gold and arsenic soil anomaly with a number of holes containing significant intercepts. A northwest trending line of drillholes covering 350 m at Lindblom Pit all have significant intercepts, and are all open to the northeast and southwest. The Hot Air Bench target has a hole with 3 m of 1.4 g/t gold and, like Goodluck, this drillhole has not been offset.

AGC has conducted intermittent surface mapping of the area surrounding the Rock Creek Deposit. AGC has also explored the property through diamond drilling and reverse circulation drilling.

Project No. 151088	Page 10-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 10-1: Exploration Targets in the Rock Creek Area (Avalon, 2002)

Project No. 151088	Page 10-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 10-1: Significant Drilling Results from Reconnaissance Exploration Prospects in
                      the TNR – NovaGold Joint Venture Area (St. George, 2000)

Hole #	Length	Au(g/t)	Length	Au(oz/t)	Hole #	Length	Au(g/t)	Length	Au(oz/t)
	(m)		(ft)			(m)		(ft)
				Bonanza	Hill
BDH90-10	3.0	2.6	10	0.075	BHR-4-08	1.5	2.5	5	0.073
BDH90-7	1.5	4.1	5	0.120	And	1.5	2.6	5	0.075
BDH90-8	1.5	2.3	5	0.066	And	3.0	3.2	10	0.093
BHC-5-01	3.0	2.0	10	0.059	BHR-4-10	25.9	2.2	85	0.064
BHR-4-01	3.0	4.1	10	0.120	And	1.5	2.4	5	0.070
And	10.7	2.6	35	0.076	And	1.5	2.8	5	0.083
And	9.1	2.0	30	0.059	BHR-6-03	1.5	25.5	5	0.744
BHR-4-02	1.5	2.6	5	0.077	BHR-6-05	1.5	5.2	5	0.152
BHR-4-04	6.1	4.5	20	0.130	BHR-6-12	1.5	8.2	5	0.239
BHR-4-05	4.6	2.5	15	0.072	BR-7-002	3.0	3.7	10	0.109
BHR-4-06	12.2	2.4	40	0.069	TDH90-05	1.5	2.0	5	0.059
BHR-4-07	4.6	2.9	15	0.086	And	1.5	3.2	5	0.094
And	9.1	3.2	30	0.093
And	1.5	3.6	5	0.105
And	3.0	2.6	10	0.077
Goodluck
GLR-6-01	3.0	20.8	10	0.608
L. Rock Creek
DH90-13	4.6	2.6	15	0.076	RDH90-14	18.3	2.3	60	0.068
Lindblom Pit
LBR-6-01	3.0	4.2	10	0.124	LBR-6-34	1.5	7.0	5	0.204
LBR-6-09	4.6	4.0	15	0.117	And	1.5	2.9	5	0.085
And	1.5	2.8	5	0.082	LBR-6-35	1.5	4.7	5	0.137
LBR-6-17	1.5	7.1	5	0.208
Saddle
SR-7-001	57.9	2.2	190	0.064	SDR-4-11	1.5	2.3	5	0.067
SDR-4-02	1.5	2.5	5	0.074	SDR-4-12	1.5	2.0	5	0.058
And	1.5	2.1	5	0.062	And	6.1	2.4	20	0.071
And	1.5	2.4	5	0.069	and	1.5	3.9	5	0.115
And	1.5	4.1	5	0.120	SDR-4-13	1.5	2.6	5	0.077
SDR-4-03	3.0	3.0	10	0.088	SR-7-002	6.1	2.1	20	0.062
SDR-4-05	3.0	2.4	10	0.071

Project No. 151088	Page 10-5
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

11.0	DRILLING
11.1	Drillhole Summary
	Table 11-1 summarizes the Rock Creek drillhole database by sample type (core and reverse circulation drilling) by company, year, and campaign. All core hole campaigns are defined by a 300-series code while reverse circulation (RC) drilling data were given a 500-series code. These codes were used extensively in subsequently described data analyses. The meterage shown is for the total length of all drillholes and includes some minor unassayed data. The average spacing for drillhole sections on the Rock Creek Deposit is approximately 30 m (see Appendix F-5), which is appropriate for classifying Indicated Resources.
	Table 11-1: Rock Creek Sampling Campaigns

		Core Holes			Year			RC Holes
Company	No.	Metres	Pre-fix	Code		Company	No.	Metres	Pre-fix	Code
	Holes						Holes
Placer Dome	24	2,033.69	RC-7	301	1987	Placer Dome	17	966.00	RR-7	501
Placer Dome	15	1,582.51	RC-8	302	1988	Placer Dome	73	4,614.00	RR-8	502
				303	1989	Placer Dome	10	532.00	RR-9	503
Tenneco	1	245.67	DDH90	304	1990	Tenneco	1	32.00		504
Newmont	6	569.49	RC92,AC92	305	1992	Newmont	1	82.00	AR-92	505
Kennecott	11	1,265.83	RCC-4	306	1994	Kennecott	21	1,584.00	RCR	506
Kennecott	20	2,464.59	RCC-5	307						507
				308	1996	Kennecott	2	305.00	BRR	508
				309						509
				310						510
				311	1999	NovaGold	6	437.40	RMR	511
				312	2000	NovaGold	30	2,444.49	RR-0	512
NovaGold	16	1,181.78	RKDC02	313						513
NovaGold	12	1,481.47	RKDC03	314						514
(spring)
NovaGold	94	10,355.65	RKDC03	315						515
(summer)
NovaGold	53	3,994.08	RKDC04	316	2004	NovaGold	27	1,938.55	RKRC04	516
	252	25,174.76			Total		188	12,935.44

Project No. 151088	Page 11-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	The drillhole prefixes can be confusing in that the Placer Dome prefix RC stands for Rock Creek Core, whereas the normal industry practice is to use the prefix RC to designate reverse circulation data.
	The 2004 AGC drilling campaign recovered nearly 4,000 m of HQ core from 53 holes. Triple-tube core recovery equipment was utilized to minimize the loss of fine-grained and broken core. Reverse circulation (RC) drilling recovered 1,900 m of sample from 27 drillholes. After photography, geologic and geotechnical logging, whole core was submitted for assay analysis in 2 m lengths. RC chip samples were submitted in 5 ft intervals. Ten trenches were also completed.
	Appendix A-13 has a list of all 5 m RC and DDH composites greater than 1.0 g/t. Some of the composite lengths are less than 5 m due to termination at the end of the hole. Some other composites may be short due to rejected samples from poor recovery or suspicious RC samples.
11.2	Review of 2004 Site Visit
	In September, 2004, Dr. Harry Parker made a site visit to the Rock Creek Property to review the drilling and trenching program (see Appendix A-12).
	AMEC had the following conclusions:
	•	AMEC was impressed with the field program and that the resultant data should support a feasibility study.
	•	Use triple-tube core barrels to improve core recovery in gold bearing soft zones that were seen in the trenching.
	•	Perform some check samples from the coarse rejects and have them assayed at another lab, as all QA/QC samples being prepared and assayed at the present lab are non-blind.
	Both diamond and RC drilling sampling practices appear to be as good as could be achieved under the observed drilling conditions.
11.2.1	RC Drilling
	The RC rig drives a 6 in casing to a depth of 30 ft. The hole is conditioned with mud and continued with a 5 ¾ in hammer using a center-return bit. In very wet conditions,

Project No. 151088	Page 11-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	the hammer breaks chips too big to get through ports in the bit. Air is lost to the formation. A tri-cone bit is used under very wet conditions; speed of drilling is reduced.
11.2.2	RC Logging
	There is insufficient time to log at the rig. Chip trays are logged later. Logging instructions and a form are reproduced in Appendix A-12. There is no flow sheet checklist for RC holes. Creating one would be useful.
	It is possible to distinguish lithology and carbonate from quartz veins in chips. Otherwise, the logging of the chip trays will not be very useful. There is simply too little sample to be very representative of a 5 ft interval.
11.2.3	Diamond Drilling
	Core quality is enhanced by use of face-discharge bits and Extreme No. 1 Granular drilling mud-polymer.
	Downhole surveys are made every 50 ft using a digital camera. Magnetic azimuths are increased by 14.3º. This was confirmed correct by AMEC using GeoMag software.
	It would add credibility, if 10 percent of the holes surveyed using GPS were also surveyed with a theodolite and the coordinates compared.
11.2.4	Coring
	HQ diametre (63 mm) cores are extracted using a 5-ft triple-tube core barrel. AMEC visited drillhole RKDC04-282. Core from a depth of about 20 ft was pulled. Soft material was recovered intact.
	Appendix A-12 includes a core-handling flow sheet. This is a checklist designed to ensure that all required activities are performed: geotechnical logging, geological logging, sample selection, photography, sampling, shipping, and data entry.
11.2.5	Core Logging
	AMEC’s site visit included a review of the core logging program. A number of log sheets are prepared for use by the Geologists.
	A core-logging manual developed by AGC is contained in Appendix A-12.

Project No. 151088	Page 11-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	The logging is of very good quality and comprehensive. AMEC spot-checked logging for hole RKDC04-271 and found it to be reasonable.
	Logs are entered using DDH Tool software. There is no duplicate data entry. This may lead to higher than desirable error rates. Best practice is to independently double-enter data.
	Graphic logs are scanned. The core is digitally photographed before sending for preparation. The quality of photographs is good.
11.3	Data Orientation
	Core, rotary and trench sample locations are found in Figure 11-1. The RC and diamond drillhole bit sizes by year and company are found in Table 11-2 (some of the RC bit sizes could not be determined). The orientation of the Rock Creek drillhole and trench data are summarized in Table 11-3. The majority of the data used for estimating mineral resources were steep easterly oriented drillholes. This orientation was chosen to intersect the steep west dipping, north-south trending shear zones.
	Table 11-2: Drillhole Sizes

Company	Year	Core Diameter	RC Diameter
Placer Dome	1987-89	HX,NX	5 in, 4.75 in
Tenneco	1990	HX	?
Newmont	1992	HQ	?
Kennecott	1994-95	HQ	5.5 in, 5.25 in
NovaGold	1999-+	HQ, HQ3	5-5.5 in

Table 11-3: Data Orientation

Drillhole/Trench Orientation	No. Holes/Trenches	No. Metres
Vertical Downward Hole	6	497.50
Steep Downward Northeasterly Angle Hole	1	64.00
Steep Downward Easterly Angle Hole	393	33,913.71
Steep Downward Southeasterly Angle Hole	1	86.11
Steep Downward Southwesterly Angle Hole	2	227.99
Steep Downward Westerly Angle Hole	26	2,489.72
Shallow Downward Easterly Angle Hole	3	179.78
Shallow Downward Westerly Angle Hole	7	732.00
Flat Easterly Trench	5	167.00
Flat Westerly Trench	3	301.30
Shallow Upward Easterly Trench	2	96.00
Total	449	38,755.11

Project No. 151088	Page 11-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 11-4: Data Inclination

Inclination	No. Holes/Trenches	No. Metres
-90	6	497.50
-80	18	1,062.41
-75	2	158.19
-70	117	7,542.49
-65	5	386.19
-60	20	1,667.32
-55	103	10,160.03
-50	15	2,012.04
-45	144	13,880.64
-15	4	319.00
-10	3	305.00
-5	4	323.30
0	6	345.00
5	1	45.00
10	1	51.00
Total	449	38,755.11

Project No. 151088	Page 11-5
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 11-1: Drillhole and Trench Location Map (5 m topographic contour interval) ¹

______________________________
1 Local Grid is rotated -50 degrees from UTM East (NAD 27) 479443 and UTM North (NAD 27) 7165469 to local grid at 0 East and 0 North.

Project No. 151088	Page 11-6
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Most of the sample data were inclined at -45 and -70 degrees to the east in an effort to intersect the steep west dipping mineralized zones.

The RC holes tended to be steeper, as drilling conditions proved difficult when trying to drill at a flatter angle. Note that the meterages shown in Table 11-3 through Table 11-5 include unassayed intervals. Table 11-3 summarizes the dip orientation of the Rock Creek sample data. Table 11-4 summarizes total meterages of all drillholes by various inclinations.

The drill holes intersect the veins at about 45 degrees and thus gives a thickness of about 140% of the true thickness for mineralized sample intervals.

Table 11-5 breaks down the Rock Creek sample data by type and inclination.

Table 11-5: Data Inclination by Type

Data Type	Inclination	No. Metres
Core	-90	183.00
	-80	234.07
	-75	81.99
	-70	1,799.68
	-65	322.18
	-60	1,522.50
	-55	7,265.95
	-50	2,012.04
	-45	11,570.70
Core Total		24,992.11
RC	-90	314.50
	-80	828.34
	-75	76.20
	-70	5,742.81
	-65	64.01
	-60	144.82
	-55	2,894.08
	-45	2,245.94
RC Total		12,310.70
Trench	-45	64.00
	-15	319.00
	-10	305.00
	-5	323.30
	0	345.00
	5	45.00
	10	51.00
Trench Total		1,452.30
Grand Total		38,755.11

Project No. 151088	Page 11-7
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

11.4	Data Summaries by Sample Campaign
	Table 11-6 summarizes the final unadjusted assay data above a 0.0 g/t gold cutoff grade by sampling campaign that was used for estimating mineral resources.
	Core hole data represent about 65% of the total sample data, yet only contain about 54% of the total grade-thickness product. The RC data represent about 31% of the data and contain about 43% of the grade-thickness product. The trench data were used to estimate mineral resources, but their area of influence was restricted to 15 x 15 x 5 m from each sample. This permits the trench data to contribute to the estimates of surficial blocks near the trenches. It is inadvisable to use the trench data unrestricted given the unexplained differences between trench assays and both adjacent core and rotary assays.
	Table 11-6: Assays by Sample Campaign (rejected samples omitted)

Campaign	No.	No.	Min	Mean	Max	Std.	CV	GT*	% of
	Samples	Metres	Au	Au	Au	Dev.			Total
			(g/t)	(g/t)	(g/t)				GT
301 - Core	1,181	1,853.70	0.01	0.48	20.06	1.31	2.71	895	3%
302 - Core	1,018	1,579.47	0.03	0.34	47.14	1.81	5.38	531	2%
304 - Core	150	241.40	0.00	0.12	3.33	0.32	2.54	30	0%
305 - Core	355	540.70	0.00	0.59	49.78	2.93	4.93	321	1%
306 - Core	829	1,262.79	0.00	0.77	33.40	2.19	2.84	975	4%
307 - Core	1,561	2,367.26	0.00	0.46	75.84	2.44	5.26	1,098	4%
313 - Core	553	1,034.16	0.01	0.81	23.31	2.31	2.84	841	3%
314 - Core	787	1,475.31	0.00	0.51	46.60	2.15	4.25	747	3%
315 - Core	5,407	10,342.51	0.03	0.60	59.60	2.06	3.44	6,194	24%
316 - Core	2,104	3,992.08	0.03	0.61	57.80	2.08	3.42	2,424	9%
Sub-total Core	13,945	24,689.38	0.00	0.57	75.84	2.08	3.65	14,055	54%
501 - RC	588	895.42	0.07	1.76	53.69	4.46	2.54	1,574	6%
502 - RC	2,871	4,458.39	0.03	0.92	117.60	3.89	4.23	4,101	16%
503 - RC	321	503.82	0.03	1.38	68.88	5.76	4.16	697	3%
505 - RC	52	79.25	0.00	0.30	2.40	0.58	1.92	24	0%
506 - RC	979	1,491.99	0.00	0.30	17.42	1.07	3.56	450	2%
511 - RC	272	414.52	0.03	1.19	16.42	1.93	1.63	491	2%
512 - RC	1,461	2,229.54	0.00	0.90	39.04	2.68	2.98	2,011	8%
516 - RC	1,241	1,891.31	0.03	0.95	55.30	2.42	2.56	1,791	7%
Sub-total RC	7,785	11,964.24	0.00	0.93	111.60	3.34	3.59	11,139	43%
713 - Trench	370	406.00	0.00	0.59	24.99	1.98	3.39	237	1%
716 - Trench	538	999.30	0.03	0.45	13.90	1.13	2.53	445	2%
Sub-total	908	1,405.30	0.00	0.49	24.99	1.43	2.95	682	3%
Trench
Grand Total	22,638	38,058.92	0.00	0.68	117.6	2.54	3.73	25,876	100%

*GT=grade thickness (metres x mean Au g/t)

Project No. 151088	Page 11-8
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

12.0 SAMPLING METHOD AND APPROACH

12.1	Introduction
	Where multiple sampling methods are used, it is common to compare the frequency distributions of one method versus another to check for biases. At Rock Creek there are three sampling methods: DDH, RC and trench. There are also various campaigns, which have used different equipment and sampling/assaying protocols (see Appendix A-3).
	The accepted assay intervals (greater than 60% core recovery and not on the RC reject list – Appendix A-11) were used for this study. A list of all composites greater than 1 g/t is in Appendix A-13.
	There is uneven spatial distribution to the locations of holes drilled with various methods and campaigns. To compare two methods, a program was written that finds 10 × 5 × 5 m blocks located within 30 m of accepted assay intervals for both methods. Then assays from each method were assigned declustering weights as described in the next section.
	Diamond drill and RC sample collection at the Rock Creek Deposit commenced in 1987 with Placer Dome. There are very little data available as to how on site sample collection was performed prior to NovaGold acquiring the property in 1999.
	Details regarding location of holes, number nature, type, and drill density are discussed in Section 11.
12.2	RC Sampling
	Reverse circulation drill samples from the 1999 and 2000 Rock Creek program were logged on-site by NovaGold personnel. A NovaGold geologist was present at all times during the drilling program and managed all sampling conducted in 1999 and 2000. Samples were collected on 5 ft intervals and consisted of a 40% split for assay purposes and a 60% split for future metallurgical or other purposes. Samples were then bagged and secured on-site to prevent purposeful or inadvertent contamination. Original assay pulps and rejects have been discarded, the remaining 60% split of each interval is stored in NovaGold’s Nome warehouse facilities (Avalon, 2002).

Project No. 151088	Page 12-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	During the 2004 site visit, the following describes the RC sampling process. The sample interval is 5 ft. The sample consists of a ¼ split (from a rotary splitter); this is caught in three 5-gallon buckets that cascade into each other. The buckets contain suspended short lengths of 6 in PVC that reduce turbulence in the bucket. Occasionally, a fourth bucket is required.
	When the interval is completed, the buckets are decanted by pouring the water out on the ground. The sample, weighing about 12 kg, is transferred to cloth bags, which can further drain while sitting on the ground by the rig.
	A 2 kg sample is put in a strainer and panned. This sample is used for logging.
	The dry weight of the sample is recorded at the preparation laboratory. The fraction split at the rig is also recorded.
	Rods are 15 ft long. A mixture of bentonite and water is injected. If a switch to a tri- cone bit is made, the first rod is 10 ft long; thereafter 15 ft rods are used.
12.3	RC Recovery and Data Selection
	AMEC was able to estimate the recovery percentage for the 2004 AGC RC drilling campaign using dry sample weights recorded by Chemex Labs similar to the method that was used for determining core recovery using the weight method. In the 2004 RC drilling program, AGC collected an approximate ¼ split of wet RC drill cuttings for each 5 ft (1.52 m) long interval that was sampled. These ¼ splits were then dried and weighed by Chemex. A theoretical dry sample weight was calculated for the 2004 RC assay intervals using the bit diametre, sampled length, and default dry density of 2.71 g/cm3. The theoretical dry weight was divided by four and used as the denominator in a recovery calculation that used the actual reported dry weight as the numerator. AMEC acknowledges that the RC recoveries that were estimated using this method are subject to some uncertainty, but submits that in general, this method provides an indication of suspect RC intervals as defined by very low or high sample weights. Gold grades are compared with RC recovery for all of the 2004 RC samples in Figure 12.1. Unlike the graph of core recovery versus gold grade as shown in Figure 12-3, RC gold grades display more of a bell-shaped curve function relative to recovery (Figure 12-1). Most of the extremely low and high RC recovery intervals tend to have low gold grades.

Project No. 151088	Page 12-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

RC recoveries are compared against down-hole depth in Figure 12-2 for the 2004 drilling data. The data shows that in general, both the low and high recovery intervals are found at shallow depths, typically within 20 m of the surface.

Based on AMEC’s and AGC’s review of the 2004 RC recovery, data intervals with less than 30% or greater than 120% RC recovery were rejected for use in Mineral Resource estimates. John Odden of AGC reviewed drill log descriptions for each low and high RC recovery interval and was able to confirm that drilling conditions may have been responsible for the suspect weights associated with those intervals.

Figure 12-1: RC Recovery vs. Gold Grade

Figure 12-2: RC Recovery vs. Hole Depth

Project No. 151088	Page 12-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

12.4	Core Sampling
	Sampling is completed on fixed 2 m intervals. This is acceptable, given the deposit will be bulk-mined with open-pit methods. Typically samples weigh 10 to12 kg.
	The whole core is shipped to ALS-Chemex in Fairbanks for sample preparation. Assaying is done in Vancouver.
	To the extent possible, sludge samples are taken every 5 ft at the rig.
12.5	Core Recovery and Data Selection
	Percent core recovery was calculated by AGC and the previous companies that completed diamond core holes at Rock Creek by dividing the length of core actually recovered from the core barrel found between “run blocks” by the drilled length as defined by the distance values posted on the run blocks. During a review of core box photos, AMEC noticed some discrepancies between the recorded core recovery and the apparent recovery by studying the photographs. A decision was made by both AGC and AMEC that where possible, a weight method would be used to define core recovery. Entire dry sample weights were available from the Chemex laboratory for all of the 2003 and 2004 core hole assay intervals due to the sampling protocols developed by AGC. A theoretical dry weight was calculated for each 2003 and 2004 core interval using the inside bit diameter, the length of the drilled interval, and a default density of 2.71 g/cm3. The actual reported dry weight for each assay interval was then divided by the theoretical weight and then multiplied by 100, which yielded the percent of material recovered. Final core recoveries for the Rock Creek core holes were determined using a prioritized system. The recovery by weight method took precedence over core recovery by the logging method where core weights were available.
	Figure 12-3 shows unadjusted gold assays plotted versus core recovery. This relationship shows that very few high-grade samples are found below 60% recovery, suggesting a possible bias in grade with respect to core recovery. Core samples with less than 60% core recovery represent about 5.7% of the total core data.
	Figure 12-4 graphically shows the recovery versus the downhole depth suggesting that increased recovery occurs with as depth increases.

Project No. 151088	Page 12-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 12-5 compares recovery with hole depth for holes drilled pre-2003 to holes from 2003 and 2004. This graph shows that the most recent holes (2004) have had the poorest recovery of the three data sets.

Project No. 151088	Page 12-5
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 12-3: Core Recovery vs. Unadjusted Gold Grade

Table 12-1: Core Recovery Statistics

Core Recovery	Incremental	Cumulative	% of	Mean Au	Max	Std.	CV
(%)	Metres	Metres	Total	(g/t)	Au (g/t)	Dev.
100 & above	9,461	23,927	39.50	0.52	30.00	1.67	3.22
90 to 99	7,605	14,466	31.80	0.54	30.00	1.70	3.18
80 to 90	3,386	6,861	14.20	0.56	23.70	1.52	2.75
70 to 80	1,260	3,475	5.30	0.58	30.00	1.65	2.85
60 to 70	672	2,215	2.80	0.70	30.00	2.42	3.44
50 to 60	417	1,543	1.70	0.74	30.00	2.38	3.23
40 to 50	295	1,126	1.20	0.60	12.00	1.43	2.39
30 to 40	326	831	1.40	0.66	30.00	2.45	3.71
20 to 30	233	505	1	0.28	5.56	0.50	1.80
10 to 20	168	272	0.70	0.79	6.45	1.41	1.80
0 to 10	105	105	0.40	0.31	6.45	0.92	3.01
Total	23,927		100	0.54	30.00	1.70	3.13

Gold grade statistics were tabulated for all of the Rock Creek core data using 10% core recovery bins as shown in Table 12-1. A cap grade of 30 g/t gold was used to reduce the affect of outlier values. Notice that in general, the CV is generally lower for core recoveries less than 60%.

Table 12-2 tabulates basic descriptive statistics by core recovery (less than and greater than 60%) as a function of distance from surfaces and by drill campaign.

Project No. 151088	Page 12-6
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 12-4: Figure: Core Recovery Statistics

Figure 12-5: Core Recovery vs. Down-hole Depth (Moving Average)

Project No. 151088	Page 12-7
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 12-2: Subdivided Core Recovery Statistics

Category	Parameter	Below 60% Recovery					Above 60% Recovery
		Metres	% of	Au	Std.	CV	Metres	% of	Au	Std.	CV
			Total	(g/t)	Dev.			Total	(g/t)	Dev.
Distance	0 to 25m	1,140	14	0.59	2.62	4.46	7,009	86	0.53	1.46	2.77
from	25 to 50m	163	2	0.79	1.61	2.03	6,489	98	0.65	2.21	3.42
Topography	50 to 75m	83	2	0.73	1.32	1.81	4,681	98	0.56	1.95	3.46
	75 to 100m	34	1	0.15	0.21	1.45	2,446	99	0.57	1.70	3.00
	+ 20 to 30m	24	4	0.62	1.21	1.94	590	96	0.63	1.22	1.94
Distance	+ 10 to 20m	47	8	0.55	1.21	2.20	540	92	0.66	1.50	2.26
from Sophie	+ 0 to 10m	89	16	0.39	0.58	1.51	482	84	0.56	1.73	3.08
Gulch Fault	- 0 to 10m	59	11	2.98	10.52	3.53	488	89	0.37	0.58	1.57
Zone	-10 to 20m	11	3	2.82	3.26	1.16	405	97	0.55	0.98	1.79
	- 20 to 30m	13	4	0.07	0.14	1.94	319	96	0.25	0.58	2.31
	1987 Placer Dome	232	12	0.48	1.47	3.08	1,641	88	0.46	1.70	3.74
	1988 Placer Dome	83	5	0.51	2.35	4.65	1,482	95	0.33	2.32	7.00
	1990 Tenneco	46	19	0.14	0.20	1.43	195	81	0.12	0.42	3.51
	1992 Newmont	6	2	1.09	1.61	1.48	369	98	0.65	1.44	2.21
	1994 Kennecott	36	4	0.42	0.73	1.75	848	96	0.58	1.60	2.74
Drilling	1995 Kennecott	87	4	0.31	0.65	2.08	2,269	96	0.47	2.45	5.20
Campaign	2002 NovaGold	44	4	1.00	1.55	1.55	1,008	96	0.77	1.97	2.55
	2003 NovaGold-	66	4	0.35	0.39	1.13	1,410	96	0.51	1.85	3.60
	Spring
	2003 NovaGold-	473	5	0.70	3.49	5.00	9,642	95	0.60	1.67	2.80
	Summer
	2004 NovaGold	369	9	0.71	1.56	2.20	3,623	91	0.59	1.66	2.80

As the data in the figures and tables above show, the number of intervals with less than 60% core recovery declines with depth. Also, the number of intervals with less than 60% recovery increase with proximity to the Sophie Gulch Fault where gouge and broken rock were noted in drill logs. The 1987 Placer Dome and 1990 Tenneco drilling campaigns seem to have an unusually high percentage of intervals with less than 60% core recovery.

Down-hole variograms were examined by AMEC for various core recovery thresholds in the spring of 2004 (Appendix A-9). It was noted earlier that the nugget effect decreased as core recovery decreased, which suggested that some coarse gold may have been lost in core intervals exhibiting low recovery. It was also noted that the down-hole variogram range increased with core recovery, indicating slightly better spatial correlation with improved core recovery.

In conclusion, AMEC determined that it was appropriate to reject core hole assay intervals with less than 60% recovery. As a result, 1,433 m or 5.7% of the core assay data were excluded from being used to estimate mineral resources.

Project No. 151088	Page 12-8
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

12.6	Review of Twin Hole Data Selection
	Twin holes are often used to compare various drilling/sampling methods and to establish local variability in grade associated with the different methods. At Rock Creek, there is an unprecedented number of twin hole data. Using methods that were established for the July 2004 AMEC report, a total of 132 twin hole comparisons were made. These comparisons included core versus RC, core versus core, and RC versus RC for a variety of historical drilling campaigns that have been completed at Rock Creek. The primary focus of AGC’s 2004 core and RC drilling programs was to determine whether previously suspected biases in sample data could be confirmed with best practice drilling and sampling procedures. Triple tube core barrels were used in the 2004 drilling campaign in an attempt to minimize suspected gold loss. The 2004 RC program utilized face discharge hammers and a mud stabilization program to minimize down-hole contamination and ravelling. The 2004 drillholes were sited at a number of previous drillhole locations. Figure 12-6 is a plan map that shows the location of the 30 twin hole drill sites that were examined for this report.
	A total of four drillholes were located at most of the twin site locations. This allowed AMEC to create six paired combinations from each of those sites. Forty-one of these twin holes were drilled in 2004 by AGC (21 core and 20 RC holes).
	The twin hole pairs were graded (scored) A through F by how well the twin hole grades matched one another. Holes that were scored A tend to have cumulative grade- thickness products with similar slopes. Given the fact that vein intercepts can vary in the down-hole location of two adjacent holes even if they are separated by only a few metres, the correlation of “spikes” in the grade-thickness profiles was not given much weight.
	The width of higher-grade zones was considered in the scoring process. Holes that were scored as B-type holes tend to show some divergence in the slopes of the grade- thickness plots. The C scored holes tend to show marked divergence of grade- thickness slopes even in low-grade material. Holes that were scored as D and F show very poor correlation in mineralized zones with large differences in the grade-thickness plots.
	Table 12-3 summarizes the drillholes that were examined for each twin hole drill site.

Project No. 151088	Page 12-9
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 12-6: Twin Hole Drill Sites ²

______________________________
2 Local Grid is rotated -50 degrees from UTM East (NAD 27) 479443 and UTM North (NAD 27) 7165469 to local grid at 0 East and 0 North.

Project No. 151088	Page 12-10
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 12-3: Twin Hole Data

Site	Approximate Location			Neighboring Holes
	East	North	Elev	Hole 1	Hole 2	Hole 3	Hole 4
1	395	337	71	RC-7-006	RKDC03-212	RKDC04-242	RKRC04-040
2	441	361	69	RR-0-03	RKDC03-209	RKDC04-237	RKRC04-023
3	416	364	69	RR-0-02	RKDC03-222	RKDC04-236	RKRC04-024
4	506	377	75	RCE-4-19	RKDC03-208	RKDC04-239	RKRC04-39
5	485	403	74	RR-7-017	RKDC03-206	RKDC04-238	RKRC04-022
6	368	454	73	RC-7-005	RMR-6	RKDC04-241	RKRC04-025
7	508	454	87	RR-0-11	RKDC03-197	RKDC04-240	RKRC04-041
8	466	455	87	RR-8-028	RKDC03-199
9	460	459	77	RMR-5	RC-7-002	RKDC04-257	RKRC04-047
10	443	470	77	RCC-4-01	RCR-4-22	RKDC04-256	RKRC04-038
11	371	519	79	RC-7-009	RKDC03-203	RKDC04-235	RKRC04-036
12	366	538	80	RR-0-15	RKDC03-202	RKDC04-434	RKRC04-035
13	381	593	83	RC-7-008	RCC-5-20	RKDC04-243	RKRC04-037
14	440	595	96	RR-8-037	RKDC03-195	RKDC04-233	RKRC04-026
15	323	658	89	RKDC03-138	RKDC04-250
16	498	659	111	RR-8-078	RKDC03-180	RKDC04-232	RKRC04-027
17	486	721	116	RR-0-28	RKDC03-193	RKDC04-231	RKRC04-028
18	353	722	99	RR-0-37	RKDC03-225	RKDC04-245	RKRC04-034
19	363	769	107	RC-7-016	RCC-4-11	RC92-003
20	366	769	105	RR-8-024	RKDC03-177	RKDC04-244	RKRC04-032
21	398	770	109	RR-0-027	RKDC03-178	RKDC04-248	RKRC04-033
22	386	815	109	RR-0-32	RKDC03-224	RKDC04-247	RKRC04-031
23	350	828	101	RR-8-079	RKDC03-179	RKDC04-246	RKRC04-029
24	488	386	75	RR-9-092	RKDC03-200
25	369	394	73	RR-7-013	RKDC03-205
26	406	414	72	RMR-4	RKDC03-204
27	401	512	80	RR-7-002	RC-07-021
28	481	530	94	RR-8-090	RKDC03-196
29	292	543	80	RCR-4-21	RKDC03-201
30	301	1141	120	RC-8-037	AC92-005

AMEC created an 11” x 17” plot for each of the 132 twin hole combinations (Appendix A-10). These plots were used to help assess twinned hole data pairs. Each of the plots were divided into the following three areas: (1) a side-by-side bar chart of the down-hole gold grades, (2) a plot of the down-hole horizontal separation between the two holes, and (3) a plot that compares the cumulative grade-thickness product of each hole versus down-hole depth. AMEC made hand notations on some of the plots showing zones of poor core recovery, suspected zones showing down-hole

Project No. 151088	Page 12-11
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

contamination, zones of low downhole CV, etc. Table 12-4 lists the scores that each hole was given after reviewing the twin hole profile plots.

Table 12-4: Twin Hole Scores

Site	Drillhole	Campaign	Rank 1	Rank 2	Rank 3	Final Rank
	RC-7-006	301	A	A	B+	A
1	RKDC03-212	315	A	A	B+	A
	RKDC04-242	316	A	A	A	A
	RKRC04-040	516	A	B+	B+	B
	RR-0-03	512	A	B	A	A
2	RKDC03-209	315	A	A	A	A
	RKDCO4-237	316	A	B	A	A
	RKRC04-023	516	A	A	A	A
	RR-0-02	512	A	A	B	A
3	RKDC03-222	315	A	A	B	A
	RKDC04-236	316	A	A	B	A
	RKRC04-024	516	B	B	B	B
	RCR-4-19	506	B	C	A	B
4	RKDC03-208	315	B	A	A	A
	RKDC04-239	316	A	C	C	C
	RKRC04-039	516	C	A	A	A
	RR-7-017	501	A	C	A	A
5	RKDC03-206	315	A	A	A	A
	RKDC04-238	316	A	C	B-	B
	RKRC04-022	516	B-	A	A	A
	RC-7-005	301	A	A-	A	A
6	RMR-6	511	A	A	A	A
	RKDC04-241	316	A-	A	A	A
	RKRC-4-025	516	A	A	A	A
	RR-0-11	512	B+	B	B	B
7	RKDC03-197	315	B+	A	A	A
	RKDC04-240	316	A	B	A	A
	RKRC04-041	516	A	A	B	A
8	RR-8-028	502	F			F
	RKDC03-199	315	F			F
	RMR-5	511	B	A	A	A
9	RC-7-002	301	B	A	C	B
	RKDC04-257	316	A	A	A	A
	RKRC04-047	516	A	C	A	A
	RCC-4-01	306	A	A	A	A
10	RCR-4-22	506	A	A	B	A
	RKDC04-256	316	A	A	B	A
	RKRC04-0438	516	B	A	B	B
	RC-7-009	301	A	B+	A	A
11	RKDC03-203	315	A	A	A	A
	RKDC04-235	316	A	B+	A	A
	RKRC04-036	516	A	A	A	A
	RR-0-15	512	B	B	A-	B
12	RKDC03-202	315	B	A	A	A
	RKDC04-234	316	A	B	A	A
	RKRC04-035	516	A	A	A-	A
	RC-7-008	301	B+	B+	C	B
13	RCC-5-20	307	B+	A	B+	B
	RKDC04-243	316	B+	A	B+	B
	RKRC04-037	516	B+	C	B+	B

Project No. 151088	Page 12-12
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Site	Drillhole	Campaign	Rank 1	Rank 2	Rank 3	Final Rank
	RR-8-037	502	A	A	A	A
14	RKDC03-195	315	A	A	C	A
	RKDC04-233	316	A	A	C	A
	RKRC04-026	516	C	C	A	C
15	RKDC03-138	315	B			B
	RKDC04-250	316	B			B
	RR-8-078	502	B	B	B	B
16	RKDC03-180	315	B	A	A	A
	RKDC04-232	316	A	B	B+	B
	RKRC04-027	516	B+	A	B	B
	RR-0-28	512	A	B+	A	A
17	RKDC03-193	315	A	A	B+	A
	RKDC04-231	316	A	B+	A	A
	RKRC04-028	516	A	B+	A	A
	RR-0-37	512	A	A	A	A
18	RKDC03-225	315	A	A	A	A
	RKDC04-245	316	A	A	A	A
	RKRC04-034	516	A	A	A	A
	RC-7-016	301	A	A		A
19	RCC-4-11	306	A	A		A
	RC92-003	305	A	A		A
	RR-8-024	502	A-	A	A	A
20	RKDC03-177	315	A-	A-	A	A
	TKFV04-244	316	A-	A	A-	A
	RKRC04-032	516	A-	A	A	A
	RR-8-027	502	B	C+	A-	B
21	RKDC03-178	315	B	B-	A	B
	RKDC04-248	316	B-	C+	B	B
	RKRC04-033	516	B	A	A-	A
	RR-0-32	512	A	C	B+	B
22	RKDC03-224	315	A	B	A	A
	RKDC04-247	316	B	C	C	C
	RKRC04-031	516	C	A	B+	B
	RR-8-079	502	A	A	B	A
23	RKDC03-179	315	A	A	B	A
	RKDC04-246	316	A	A	A	A
	RKRC04-029	516	A	B	B	B
24	RR-9-092	503	D			D
	RKDC03-200	315	D			D
25	RR-7-013	501	C			C
	RKDC03-205	315	C			C
26	RMR-4	511	A-			A
	RKDC03-204	315	A-			A
27	RR-7-002	501	C			C
	RC-7-021	301	C			C
28	RR-8-090	502	B			B
	RKDC03-196	315	B			B
29	RCR-4-21	506	D			D
	RKDC03-201	315	D			D
30	RC-8-037	302	A			A
	AC92-005	305	A			A

Project No. 151088	Page 12-13
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	Each hole received up to four different scores depending on how it compared with each of its neighbouring twins. A final score was given to each hole based on the majority of all scores. The scored twin drillhole data were compared in a series of quantile-quantile (Q-Q) plots and provided a benchmark as to how different sampling methods (i.e. core versus RC) compared on a campaign-by-campaign basis. These benchmark comparisons from the scored twin hole data were used to validate various adjustment factors that were developed for the entire RC data set.
12.7	Downhole Coefficient of Variation
	There are several different methods of detecting or flagging contaminated intervals in RC drillholes. The method of testing for cyclicity is a good check for ravelling of mineralized material down the drillhole during rod changes. Some companies try to use downhole variograms to detect this cyclicity. Decay analysis helps to detect asymmetric grade profiles near high grade gold spikes. AMEC attempted to use downhole correlograms, but could not see clear evidence of differences when looking at clearly contaminated holes (RR-8-028, grade = F) versus holes that correspond well with their twin.
	Figure 12-7 through Figure 12-9 contains the mean versus CV plots for drillholes in groups A, B and C respectively. The “A” drillholes tended to have higher CVs. Realizing this general trend in mean versus CV for drillholes by group3, it was decided to try using the CV to flag suspicious intervals.
	AMEC measured the CV within drillholes so that portions of drillholes could be flagged as being suspicious. This would allow resource modellers to use the CV-flagged intervals as a guide for problem areas when looking through sections for problem areas. Runs of ten samples (15 m) were used for calculating means and CVs within RC holes. This provided a better definition than whole drill-hole CVs, while still using enough data to identify less variable zones. Figure 12-10 shows the result of these calculations for the twin holes. Only Figure 12-11 showing the results for all holes. The pattern of points is very similar to that shown in Figure 12-10. The twin hole RC data are representative of the entire database. A line was fitted so that nearly all the “A” scored drillhole values were above it. A closer examination shows most of the points below the line were graded C, D, and F. This procedure provides a way of flagging some of the worst ranked sections of holes for later review. Intervals were highlighted on sections. These low CV areas were considered with other available data, and a

______________________________
3 Where contamination occurs, there is mixing of material in the sample from uphole intervals. This tends to reduce the CV. Some cognizance of local grade must be taken. In the vicinity of “nuggety” gold, the CV will increase.

Project No. 151088	Page 12-14
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

consistent method of rejection was employed. AMEC and AGC personnel agreed to all rejected intervals.

Figure 12-7: Group A, CV vs Mean Gold

Figure 12-8: Group B, CV vs Mean gold

Project No. 151088	Page 12-15
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 12-9: Group C, CV vs. Mean Gold

Figure 12-10: CV vs. Mean Gold-Twin Holes

Project No. 151088	Page 12-16
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 12-11: CV vs. Mean- All Holes

	Appendix A-11 contains a list of RC rejected intervals. Previously prepared notes by Stan Dodd and Stephen Blower are included; these were considered, but in four instances were over-ridden where new drilling supports the RC grade.
	The rejection performed here is designed to eliminate clearly suspicious samples. There is still residual bias between the RC and diamond drillhole (DDH) drilling methods.
12.8	Trenching
	Trenches appear to have been cut with a backhoe excavator, and are approximately 2 m wide and 4 m deep. The length of the trenches is several hundred metres. The primary objective of the trenches was to sample the tension shear veins.
	The decomposed rock underneath the soil in the trench walls, which was reddish brown iron oxide, was viewed for up to several metres. Panning of this material produced gold flecks, and AGC confirmed that anomalous gold grades do occur in this material. Therefore, recovery of this material in drilling should be recognized.
	It is easy to see how gold in oxidised rock could be lost using the conventional coring process in oxidized rock as the fine-grained iron oxide might not be totally recovered.

Project No. 151088	Page 12-17
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	The rock near the bottom of the trenches appeared solid. It will be difficult to argue gold loss in the coring process below depths of 5 m.
	The trench maps should be analyzed in detail. It is possible that the veins can be selectively mined, reducing dilution. With good sampling and pit mapping, the 10 X 10 X 5 m selective mining unit (SMU) now used could be reduced to perhaps 10 X 5 X 5 m. This would require angle RC grade-control drilling on 6 m spacing and use of hydraulic excavators for mining.
12.9	Sludge Sample Examination
	AGC collected and analyzed sludge samples from most 2004 core holes. AGC had not intended to use these samples for quantitative analysis or comparison to core samples (John Odden, pers. comm.). A bucket was placed at the collar of the core hole to capture rock debris laden drilling fluid. The drilling fluid and suspended rock was allowed to cascade into and out of the collection bucket; the material remaining in the bucket at the end of the drilling run was labelled and retained as the sludge sample.
	AMEC attempted to compare sludge samples to core samples to determine if alleged gold loss in core samples could be discerned and measured. Unfortunately, any gold loss would be indistinguishable from the other sources of gold in the sludge samples, gold liberated by the drill bit in the process of coring, and gold from the wall of the core hole. The volume of the sludge sample and the method of collection do not provide a credible comparison to core. AMEC created plots of all sludge samples versus core samples for holes with both samples. In general, the sludge samples confirm the presence and absence, but not the amount of gold assayed in core samples.

Project No. 151088	Page 12-18
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

13.0	SAMPLE PREPARATION, ANALYSES, AND SECURITY
13.1	Laboratory Sample Procedures
	Appendix A-3 provides a summary of laboratory sampling procedures used at Rock Creek since 1987. During the earlier years (1987-1990), some of the basic procedures have limited detail. The Rock Creek deposit has used mostly Bondar Clegg and ALS Chemex for all analytical procedures except for 1988, when Placer Dome used the Placer Dome Technical Center for both core and RC sample assaying. There is no mention of the laboratory certification in the reviewed documentation, although most assay laboratories since the early 2000s need certification to be competitive.
	On September 20, 2004, Arne Bakke CPG, former Chief Geologist for Fairbanks Gold and Kinross, made a site visit to ALS Chemex in Fairbanks Alaska to observe and comment on the sample preparation procedures that were performed on the Rock Creek samples (see Appendix A-7). His conclusions from the visit were that NovaGold is receiving quality sample preparation from ALS Chemex in Fairbanks.
	On September 22, 2004, Steven Blower, P. Geo with AMEC, made a site visit to ALS Chemex’s laboratory in Vancouver to review the metallic screen sample preparation procedures for the Rock Creek Project (see Appendix A-7).
	The initial sample preparation was performed in the ALS Chemex lab in Fairbanks and flown to Seattle and then trucked to the Vancouver Chemex lab. The samples arrived in Vancouver in multiples of 20 and each set was processed as one work order to ensure that standard reference material (SRM) controlled sets flow through the laboratory as a group. Steven Blower’s conclusions were that the metallic screen procedures being used for NovaGold met or exceeded standard industry practices.
	In January, 2005, Scott D. Long, P. Geo, with AMEC, performed a review of QA/QC samples that were assayed during the field seasons of 2003 and 2004.
	The 2003 to 2004 Rock Creek resource drilling programs include 10,331 assay results from 186 drillholes. There are 8,778 assayed intervals; the other 1,553 assay results are related to quality control, including duplicates, inserted standards and blanks (see Appendix A-5).
	A tally of Quality Control assays is shown in Table 13-1.

Project No. 151088	Page 13-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 13-1: Assay Quality Control Samples

		2003 Summer			2004 Summer
			Sample Type
	# Holes	% of Total	# holes	N	% of Total	# Holes
Routine Samples	5,106	85	94	3672	85	89
Duplicates	302	5	94	218	5	89
Field Blanks	301	5	94	218	5	89
Std-2Pa	160	3	71
Std-7Pa	140	2	68	41	1	40
Std-15Pz				84	2	69
Std-50P				93	2	73
Other				3	0.1	3
Total	6,009			4,329

	There are 302 assay certificates associated with the 2003 campaign, of which 300 include a field blank, 298 include at least one standard reference material, and 301 include at least one blind duplicate sample.
	There are 218 assay certificates associated with the 2004 campaign, of which 216 include a field blank, 216 include at least one standard reference material, and 216 include at least one blind duplicate sample.
	Quality control coverage is extensive and adequate to the needs of evaluating assay quality for use in resource models.
13.2	SGS Check Assay Program Results
	For a detailed review of QA/QC results from the 2003 to 2004 program, see Appendix A-5.
	Accuracy has been monitored by inserted geostatistical standards, blanks, and check assays. These indicate that the ALS Chemex results have a low bias, possibly due to episodic gold losses. The estimated overall low bias averages less than five percent and is therefore acceptably accurate for resource modeling.
	Precision is consistent with samples that have a sizable “nugget effect”. The effect is seen in duplicate splits of the coarse reject duplicate and in the duplicate and triplicate samples of the minus fraction of the screen fire assays. Additional assays of aliquots of the minus fraction could provide marginal improvements to precision, in cases where the duplicate or triplicate assays show high variability. Some marginal improvement might also be gained in the few cases where the plus fraction is

Project No. 151088	Page 13-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

anomalously low in mass but high in grade, as these samples may have undergone over-grinding or gold loss. While this might provide some slight improvement, the number of samples involved is too small to likely influence global resource estimates.

At AMEC’s request, AGC submitted 77 samples of 10 mesh coarse sample reject to SGS Lakefield in Canada for check screen fire assays. An analysis of the results is provided in Appendix A-5. These samples come from the 2004 drilling campaign.

A list of samples for check assaying was performed through a semi-random selection from the deposit ore zones. No original assay criterion was used in the selection process; thus the selected samples span a wide grade range and should be considered representative of those samples within mineralized drilling intercepts or down-hole drilling composites likely to be classified as ore grade. In order to obtain wide geographic coverage, the number of samples randomly selected per drillhole was restricted.

Approximately 100 samples were requested. Chemex Laboratories in Fairbanks was unable to locate some of the samples in their sample storage. A few additional samples that were recorded as shipped are absent from the SGS check assay results. A total of 77 drill samples were included in the final submission, plus some control samples including blanks and duplicates. In three cases SGS received samples weighing more than 7 kg, and these three samples were split into "A" and "B" fractions that were then assayed as two separate samples. Thus the submission includes 74 unique drill samples. The selection covers a total of 53 drillholes. There are 90 drillholes from the 2004 drilling campaign in the resource area.

The SGS check assays returned slightly higher gold results compared to the original Chemex results, with an estimated relative bias of approximately nine percent (9%). The relative bias is likely either caused by a high bias in the SGS results, a low bias in the Chemex results, or a combination of these. The standard reference material (SRM) that was inserted in the SGS sample submission appear to have probable misidentification or misstatement of best values for some samples; hence they cannot be used with confidence to estimate an absolute bias to the SGS results.

One possible explanation for a low bias at Chemex, if it exists, is the increased grinding samples underwent at Chemex, where samples were more intensely ground in order to maintain a small oversize fraction. Intense grinding can in some cases lead to gold losses as gold plates out on pulverizer surfaces. Such losses might reveal themselves in elevated gold in sample blanks. Such losses would not be uniform, but would tend to occur in samples that have large malleable gold-bearing particles.

Project No. 151088	Page 13-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	AMEC recommends conducting sampling studies to confirm and determine the source of gold loss during sample preparation.
	The Chemex results appear to provide gold results that have little risk of over- estimating gold grades.
	In the opinion of AMEC, the QA/QC results demonstrate that the Rock Creek project assay database is sufficiently accurate and precise for resource estimation
13.3	Review of Sampling Quality
	Appendix A-3 provides a description of Rock Creek sampling and assay methods, prepared by John Odden and Stan Dodd of AGC. Appendix A-4 provides a report prepared halfway through the 2003 field season by Steven Blower of AMEC. Appendix A-5, prepared by Steven Blower and Scott Long of AMEC, provides comments on QA/QC for assaying most of the latter half of the 2003 campaign and the 2004 campaign. Finally, AMEC has inserted Appendix A-7, which contains laboratory and sample preparation facility inspections by other Qualified Persons.
	A summary of these works and comments are as follows:
	•	Details for the Placer-Dome and Tenneco programs are incomplete. In general, small samples of a few hundred grams were taken for pulverization. Kennecott and Newmont took larger samples, as did AGC from 2002. Most protocols used metallic screen re-assays to replace any initial fire assay on 30 g to 50 g aliquots grading over 1 g/t. In 2003 and 2004, the 315 and 316 campaigns split and pulverized a 4 kg sample with 100% metallic screen assaying.
	•	Steven Blower’s work led to increasing the sample size and to 100% metallic screen assaying. A subsequent analysis of the RC data has been made using the twin information in January to February 2004.
	•	There was a concern that some of the intervals with low-grade fire assays that were not selected for metallic screen assay would return higher values if they had indeed been subjected to the more accurate metallic screen assay process. Scott Long concluded that any such bias (termed selection bias) would likely be small, less than 11%, and re-assaying old pulps not already assayed by metallic screen assays using the metallic screen assay technique was not warranted. This conclusion was based on a statistical analysis of limited duplicate data.

Project No. 151088	Page 13-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

14.0	DATA VERIFICATION
14.1	Drillhole Survey Check
	In October, 2004, R. Scott McClintock, PLS, was asked to re-survey drillhole locations that have been included in the resource estimation (see Appendix A-2). A total of 84 drillholes were completed in this survey with an Ashtech Z-Surveyor, dual-frequency GPS. Of the 84 holes surveyed, Scott McClintock commented that five had locations that were in question, meaning that the surveyor was not sure if he had found the actual hole location. There is no documentation as to why the locations were in question. Also, there is no documentation as to how accurate these survey co- ordinates are, as compared to the original co-ordinates of these drillholes. AMEC recommends that these holes be compared to the original survey co-ordinates to verify their location.
14.2	Database Check
	Appendix A-6, supplied by Susan Lomas, Principal Geologist with AMEC summarizes the database checks completed by AMEC dated January 8, 2004.
	The initial check of the database centered on the 2003 drilling program that included 126 drillholes and approximately 8,800 samples, including all QA/QA data. A total of  nine holes, chosen at random, were checked against all original sources including collar location, downhole survey, assay results, sample intervals, lithology code and intervals, RQD and density data. The data were found to be of excellent quality with only one data entry error.
	Another check was a comparison of the pre-2003 data in a MineSight export file to the original database previously audited by AMEC. The following discrepancies were found:
	•	A total of 279 sample assays from the 2003 database were not the same as were in the 2002 database
	•	Novagold personnel overwrote 235 of the assay results, as per AMEC’s recommendations after the metallic assays were located
	•	Another 43 assays were overwritten from the 2002 database after locating the original assay certificates

Project No. 151088	Page 14-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

•	A total of 17 sample intervals that did not match the two databases, which were subsequently corrected by NovaGold personnel
•	A total of 57 sample intervals were in the 2003 export file, but not in the 2002 export file. Also, there were 37 samples in the 2002 export file but not in the 2003 export file. AMEC recommends that NovaGold determine why these discrepancies exist in the export files.
The review of the databases identified a small number of problems, mostly related to substitution of one assay for another. It would be useful to perform a 100% check of pre-2003 data and identify the appropriate assay certificates considered to furnish “final” values for each sample interval. The QA/QC data must also be compiled in an organized group of datasets. A summary document on all QA/QC programs should be prepared. This may affect final resource classification. AMEC concludes that the assay and survey database transferred to AMEC is sufficiently free of error to be adequate for resource estimation.

Project No. 151088	Page 14-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

15.0	ADJACENT PROPERTIES
15.1	Big Hurrah Deposit
	The Big Hurrah deposit is located on the Seward Peninsula along the west coast of Alaska, north of Norton Sound. The project area lies about 80 km east of Nome and is accessed via state maintained roads.
	The terrain is fairly hilly with narrow valleys. Vegetation at the site consists mainly of low shrubs and grasses. Forested areas and trees are non-existent in the project area.
	There currently are no unusual social, political or environmental encumbrances to exploration, development or production on the prospect.
	Norwest Corporation (“Norwest”) was engaged in conjunction with Resource Modeling, Inc. (“RMI”) by NovaGold to construct a resource model for the Big Hurrah deposit in conjunction with the Rock Creek Project updated economic review study. AMEC has had no involvement in the Big Hurrah deposit and is only reporting the published information available and therefore assumes no responsibility for the validity of the data. AMEC has been unable to verify the information pertaining to the Big Hurrah deposit and such information is not indicative of the mineralization on the Rock Creek project.
	Mineralization in the Big Hurrah deposit occurs mainly in steeply dipping fissure zones that range from one to fifteen metres wide. Additional mineralization is present in low angle or bedded units in proximity to high angle structures. At the intersection of low and high angle structures the zone of mineralization may expand considerably.
	Sectional polygons of mineralization based on drillhole logging of shearing, stock work, quartz veining, and gold grade were used to construct five distinct three-dimensional units. The units trend north-south (relative to the rotated mine grid) and have varying thicknesses and dip angles in the cross strike direction. Most of the units or zones are between 5 m and 15 m thick. All of the zones dip to the west (mine grid) except Zone 4, which dips steeply to the east (mine grid).
	The five mineral zones were intersected with the northeast (mine grid) trending Hurrah and West Hill faults located at the south and north ends of the mineralized shear zones, respectively. The shear zone shapes were clipped with the faults so that the zones did not extend south of the Hurrah or north of the West Hill faults. The five main

Project No. 151088	Page 15-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

mineral zones used to constrain grade estimation are shown in a perspective view in Figure 15-1. A sixth zone labelled as 99 is made up of all material in the model falling outside of these five mineral zones. Very little “ore” grade mineralization occurs in Zone 99, and its estimated resources are classified as Inferred resources.

A large majority of the sample data used to estimate resources in the Big Hurrah deposit comes from core drilling. There are some trench samples and reverse circulation drillholes. A small number of intervals from reverse circulation drilling that appear to be contaminated with gold from further up the hole are not included in the estimation process. Based on an analysis of the frequency distribution of the sample gold values and the spatial continuity of higher grades, a cap value of 70 g/t was applied to samples. The 16 composite samples with grades greater than 32 g/t after the cap was applied to individual samples were allowed to influence grade estimates only in the block where they occurred. Estimated resources in total and by zone are given in Table 15-1.

Figure 15-1: Big Hurrah Mineral Zone Perspective

Table 15-1: Big Hurrah Mineral Resource

	Indicated Mineral Resources			Inferred Mineral Resources
Au Cutoff (g/t)	Tonnes (‘000)	Au (g/t)	Au (‘000 oz)	Tonnes (‘000)	Au (g/t)	Au (‘000 oz)
0.5	1,661	3.51	187	1,305	1.65	69
1.0	1,307	4.26	179	667	2.57	55

Project No. 151088	Page 15-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

1.5	1,062	4.96	169	427	3.33	46
2.0	882	5.62	159	289	4.10	38

Project No. 151088	Page 15-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

16.0	MINERAL PROCESSING AND METALLURGICAL TESTING
16.1	Ore Types
	The main Rock Creek ore body consists of two distinct mineralization types, Albion shear veins and tension vein ores. The Albion shear veins are best described as sheared quartz vein breccias with crushed and broken material in the shear. The tension veins are tensional quartz veins and stockwork adjacent to the shear veins. The Albion veins contain fine-grained sulfides and sulfosalts, while the tension veins have a relatively simple mineralogy of quartz, iron carbonates, sulfides and free gold. The Albion material is the most difficult ore type present at Rock Creek and drives the process scenario required at Rock Creek.
	During the course of this study, the Big Hurrah satellite deposit was reviewed and brought into the overall suite of ores that would be processed at the Rock Creek facility. The host rocks for mineralization at Big Hurrah are carbonaceous metamorphic rocks of the Nome Schist Group. Gold occurs primarily in its native state and is found in quartz veins encased in northwest striking and moderately southwest dipping thrust fault zones.
	In all of the ore types previously discussed, the vast majority of the gold present exists as free gold and is not refractory.
	The term “ore” in the context of this section, implies material delivered to a mill for processing and does not imply economic extraction or recovery.
16.2	Metallurgical Test Work
	This section has been summarized from the report submitted by Norwest Corporation (2004), as AMEC was not involved in any metallurgical testwork conducted on the Rock Creek ore and has not reviewed any of the results. Therefore, AMEC accepts no responsibility of the validity of these results.
	Metallurgical testwork on Rock Creek ore includes:
	•	Newmont Mining Company indicated greater than 80% of gold reports to a gravity concentrate with a 48 mesh grind.
	•	Placer Dome Corporation testwork indicated 92% and 93% recovery with cyanidation and flotation respectively. Surface samples for both of these studies were taken from the sheeted (tension) vein area.

Project No. 151088	Page 16-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

•	AGC completed a series of additional bench and pilot-scale metallurgical tests. The two main mineralization types were tested: Albion shear veins and tension veins. This test work showed recoveries for this material averaged 90.7% overall using cyanide, with 37.4% of the gold reporting to a gravity concentrate using a 65-mesh grind (P80 65M).
•	Earlier testwork (date not available) performed by McClelland Laboratories Incorporated in Reno, Nevada, focused on treating the whole ore with cyanidation and gravity followed by whole-ore cyanide leaching of the gravity tails. This work clearly indicated that the Rock Creek ores were amenable to gravity and cyanide leaching.
•	The most recent test-work program focused on obtaining sufficient test work data to develop a process flow sheet that would have the highest financial return for Rock Creek ores. Big Hurrah ores were also sampled and tested in this program, as AGC plans to mill this ore at the Rock Creek mill facilities.
•	High power costs in Nome, coupled with high freight costs directed the testing program to examine minimized grinding requirements and reagent needs. A test work flow sheet was developed to focus on a coarse grind gravity and flash flotation circuit that would be followed by on-site concentrate treatment. The concentrate treatment test work focused on methods to remove the gold from the concentrates in order to produce a doré product suitable for shipment to a precious metals refinery. This most recent program was performed by Process Research Associates in Vancouver, British Columbia, and Resource Development Inc. in Wheat Ridge, Colourado.
•	Test work on all ore types show a high recovery can be obtained by using a combination of gravity concentration and flash flotation. The gravity middlings and the flotation concentrate can then be effectively leached using a weak cyanide solution in a 40% to 50% solids slurry. Due to the presence of organic carbon in the Big Hurrah ore, the cyanide leaching for this ore type was observed to give highest recovery when leaching occurred in the presence of activated carbon.
Table 16.1 summarizes the results of the test work performed on the three ore types when the gravity-flotation-cyanide leach circuit was utilized.

Project No. 151088	Page 16-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 16-1: Gold Recovery by Designated Metallurgical Process at a P80 of 212 microns
                      (Norwest Corp, 2004)

Ore Type	Gravity Gold	Flotation Gold	Leach Gold	Combined Process Gold
	Recovery	Recovery	Recovery	Recovery
Tension Vein	93.0	71.6	74.9	96.8
Albion Shear	54.5	55.5	87.3	76.5
Big Hurrah	75.8	38.4	92.4	84.4

	Gold recovery was found to be optimal on the tension vein material at a grind of 212 microns. Since the tension vein material makes up approximately 75% of the ores treated at Rock Creek, this became the target grind for future test work. However, test work results on the Albion shear zone ore, were slightly lower than predicted at this target grind. Gold recoveries for the Albion shear zone material responded well at 212 micron, but were somewhat improved at finer grinds (81.5% versus 76.5% recovery).
16.3	Ore Processing
	Early in the design phase of the project, an economic analysis was performed to determine the optimum milling rate (throughput) at Rock Creek. This study was performed knowing that a second study would likely be necessary at or near the conclusion of the feasibility study, when more accurate mining and metallurgical data became available. This initial study had a throughput rate of 7,000 t/d, which optimized the economics of the project. The throughput rate appeared to be more a function of the average head grade, than any other parameter.
	Since tension vein ores are the majority of the ores processed at Rock Creek, and since gold recovery was found to be optimal on the tension vein material at a grind of 212 microns, the proposed process plant designed for the initial study was set to meet this target grind. This target grind also minimizes the power requirements for grinding. The process plant includes a gravity circuit to capture coarse gold, flash flotation cells to capture the finer gold, and a concentrate leach circuit to produce a doré bar on site. A simplified flow sheet and material balance depicting the proposed process is shown in Figure 16-1. The goal of minimizing reagent consumption is met by designing the process to only leach the concentrates, which constitute approximately 15% of the total ore stream.

Project No. 151088	Page 16-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 16-1: Process Flow Diagram

Project No. 151088	Page 16-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

17.0	MINERAL RESOURCE AND MINERAL RESERVE ESTIMATES
17.1	Gold Model Definition
	The mineral resource estimate for the Rock Creek deposit was completed under the direction of Dr. Harry Parker, P. Geo. The estimates were made from 3D block models utilizing commercial mine planning software (MineSight®). Block size for the resource model was 10 m north-south x 5 m east-west x 5 m elevation.
	The Rock Creek model is oriented parallel to a local mine grid coordinate system that is in turn oriented along the strike of the Rock Creek shear zones. Table 17-1 summarizes the limits and block dimensions of the resource model.
	Table 17-1: Resource Model Limits

Parameter	Min	Max	Extent (m)	Size (m)	Number
Easting	0	1000	1000	5	200
Northing	-100	1500	1600	10	160
Elevation	-100	200	300	5	60

Aerial topography coordinates were obtained in UTM (NAD 83) units. These coordinates and all other project data are rotated and translated into the mine grid using the following equations:

	•	Local Easting: UTM Easting – 479443 * COS(-50°) rotation + UTM Northing – 7165469 * SIN(-50°)
	•	Local Northing: UTM Northing – 7165469 * COS(-50°) rotation - UTM Easting – 479443 * SIN(-50°)
17.2	Drill-Hole Data Preparation
	A number of RC sample intervals were previously identified by AMEC as being potentially contaminated, and the intervals were rejected for use in estimating mineral resources. All of these rejected intervals were compared with AGC’s 2004 core hole data. Some of the initially rejected intervals were supported by core data and subsequently taken off the rejection list. The final pre-2004 RC rejected samples are listed in an ASCII file named dat202.rj1 and are summarized in Appendix A-11. Some 2004 AGC RC samples were rejected by AMEC based on low down-hole CVs and either low or heavy sample weights. These rejected intervals are listed in an ASCII file

Project No. 151088	Page 17-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

called dat202.rj2 and are summarized in Appendix E-1. An integer code of 9 was written to the assay variable REJEC for all RC intervals that were rejected. Core-hole intervals having a recovery less than 60% were rejected because it is the opinion of AMEC that assays from these intervals are not representative.

The raw assay gold grade variable AUORG was copied to a spare variable named AUREJ. The AUREJ variable was then set to -1.000 for all RC intervals that were rejected (i.e. REJEC = 9) and all low core recovery intervals.

The gold assay variable AUREJ was then copied to two other grade variables, AUAJ1 and AUAJ2. RC gold grade formulas were applied to the variable AUAJ1 based upon their drilling campaign code. Positive grade adjustments were made to the variable AUAJ2 for various core campaigns. The adjustments are explained in Section 17.12.

The raw drillhole data were coded with different Mineral Zone, bedrock, and fault surfaces. Table 17-2 summarizes key MineSight® wireframes that were used for coding assay, composite, and block model items.

Table 17-2: MineSight^® Geologic Objects

Assay, Composite,	MineSight	Assay,	Description
Model Variable	Object Name	Composite, Model
		Code
MINZN	Zone 1 Solid	1	Mineral Zone 1
	Zone 2 Solid	2	Mineral Zone 2
	Zone 3 Solid	3	Mineral Zone 3
	Albon Solid	10	Mineral Zone 4
BRDPT	BR-000-025	1	0-25m below bedrock surface
	BR-025-050	2	25-50m below bedrock surface
	BR-050-075	3	50-75m below bedrock surface
	BR-075-100	4	75-100m below bedrock surface
SGFLT	SG1	1	0-10m above Sophine Gulch Fault
	SG2	2	10-20m above Sophine Gulch Fault
	SG3	3	20-30m above Sophine Gulch Fault
	SG4	4	0-10m below Sophine Gulch Fault
	SG5	5	10-20m below Sophine Gulch Fault
	SG6	6	20-30m below Sophine Gulch Fault

Drillhole assays (AUORG, AUREJ, AUAJ1, AUAJ2) were composited into 5 m down-hole fixed-length composites. After the composites were created, they were coded with the Mineral Zone, bedrock depth, and Sophie Gulch Fault wireframes named in Table 17-2. A 50% majority rule was used in coding the composites.

Project No. 151088	Page 17-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

17.3	Unadjusted Assay Statistics
	Basic descriptive statistics are summarized in for the final Rock Creek drillhole database by mineral zone at four gold cutoffs. The data shown in Table 17-3 were not declustered and represent unadjusted data with all rejected intervals included. See Appendices C-1 and C-2 for histograms and probability plots of assay and composites, respectively.
	Table 17-3: Unadjusted Gold Assays by Mineral Zone

Mineral Zone		Uncapped Statistics Above Cutoff - No rejections or Adjustments
	Cutoff	Total	Inc.	Mean Au	grd-thk	Inc.	Std.	CV
	(g/t)	Metres	Percent	(g/t)	(g/t-m)	Percent	Dev.
All	0.00	38,059	76%	0.68	25,860	14.0%	2.54	3.73
	0.50	9,312	11%	2.39	22,236	11.20%	4.73	1.98
	1.00	5,192	7%	3.72	19,331	14.1%	6.00	1.61
	2.00	2,586	7%	6.06	15,676	60.60%	7.84	1.29
1	0.00	10,258	60%	1.07	10,988	10.7%	3.08	2.88
	0.50	4,100	18%	2.39	9,816	11.7%	4.56	1.91
	1.00	2,290	11%	3.73	8,533	14.4%	5.76	1.55
	2.00	1,162	11%	5.99	6,955	63.3%	7.42	1.24
2	0.00	7,128	77%	0.53	3,780	21.3%	1.61	3.04
	0.50	1,667	12%	1.78	2,975	15.2%	3.01	1.68
	1.00	844	7%	2.84	2,399	18.5%	3.94	1.39
	2.00	349	5%	4.87	1,700	45.0%	5.52	1.13
3	0.00	1,187	74%	0.97	1,156	7.8%	5.41	5.56
	0.50	312	11%	3.41	1,066	7.9%	10.16	2.97
	1.00	185	7%	5.27	975	11.1%	12.87	2.44
	2.00	100	8%	8.50	846	73.2%	16.89	1.99
10	0.00	4,165	54%	1.55	6,461	6.1%	4.10	2.64
	0.50	1,918	16%	3.16	6,065	7.1%	5.63	1.78
	1.00	1,258	13%	4.46	5,605	11.7%	6.59	1.48
	2.00	718	17%	6.75	4,847	75.0%	7.99	1.18
99	0.00	15,321	91%	0.23	3,474	33.4%	1.01	4.47
	0.50	1,314	5%	1.76	2,314	14.2%	3.04	1.73
	1.00	615	2%	2.96	1,820	14.2%	4.13	1.40
	2.00	258	2%	5.15	1,328	38.2%	5.69	1.11

Inc = Incremental Percentage
grd-thk = grade-thickness

17.4

Adjusted Assay Statistics

Basic descriptive statistics are summarized in Table 17-4 for the final adjusted Rock Creek drillhole database by mineral zone at four cutoffs. The data were not declustered and represent the final accepted assay data that were used for grade estimation. Suspicious RC samples and core samples with less than 60% core recovery are not included. RC data were adjusted to core data by drill campaign as explained in Section 17.12.

Project No. 151088	Page 17-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 17-4: Adjusted Gold Assays by Mineral Zone

Mineral Zone	Cutoff (g/t)	Total metres	Inc. Percent	Mean Au (g/t)	grd-thk (g/t-m)	Inc. Percent	Std. Dey	CV
Totals	0.00	36,424	77%	0.60	21,898	15.80%	2.12	3.53
	0.50	8,263	10%	2.23	18,432	12.2%	4.04	1.81
	1.00	4,471	6%	3.53	15,760	14.8%	5.15	1.46
	2.00	2,162	6%	5.79	12,519	57.2%	6.70	1.16
1	0.00	9,836	64%	0.87	8,572	13.7%	2.19	2.51
	0.50	3,546	17%	2.09	7,400	13.4%	3.31	1.59
	1.00	1,917	10%	3.26	6,249	16.8%	4.15	1.27
	2.00	893	9%	5.38	4,808	56.1%	5.33	0.99
2	0.00	6,840	79%	0.48	3,276	23.50%	1.56	3.25
	0.50	1,452	11%	1.73	2,506	16.8%	3.07	1.78
	1.00	677	6%	2.89	1,957	16.9%	4.20	1.45
	2.00	288	4%	4.88	1,404	42.8%	5.87	1.20
3	0.00	1,135	76%	0.71	812	10.1%	3.50	4.90
	0.50	273	11%	2.67	730	10.0%	6.77	2.53
	1.00	152	8%	4.25	648	15.1%	8.75	2.06
	2.00	66	6%	7.99	525	64.7%	12.36	1.55
10	0.00	4,016	55%	1.56	6,250	6.2%	4.15	2.67
	0.50	1,815	15%	3.23	5,863	6.8%	5.75	1.78
	1.00	1,202	13%	4.53	5,438	11.4%	6.71	1.48
	2.00	695	17%	6.80	4,726	75.6%	8.09	1.19
99	0.00	14,597	92%	0.20	2,988	35.3%	0.83	4.06
	0.50	1,177	4%	1.64	1,934	15.5%	2.49	1.52
	1.00	522	2%	2.81	1,469	13.8%	3.39	1.21
	2.00	221	2%	4.78	1,056	35.3%	4.51	0.94

17.5	Adjusted Composite Statistics
	Assays adjusted to core with suspicious RC samples and core samples with less than 60% recovery removed were downhole composited monotonically into 5 m composites. Basic declustered descriptive statistics are summarized in Table 17-5 for the final adjusted Rock Creek drillhole database by mineral zone. Note the large reduction in CV in Mineral Zone 99 once it is partitioned into two domains. See Appendix A-13 for all drillhole composites greater than 1.0 g/t.
	Table 17-5: Adjusted Declustered Gold 5 m Composites by Mineral Zone

Minzone	Number of Composites	Mean Au g/t	CV
1	2083	0.872	1.53
2	1467	0.423	2.22
3	237	0.597	3.43
10	838	1.373	1.95
99	3188	0.182	2.45
99 <0.4 g/t	2830	0.086	1.05
99 >0.4 g/t	358	1.030	1.01

Project No. 151088	Page 17-4
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

17.6	Contact Plots
	Contact plots show average grade profiles in the vicinity of contacts between mineral zones. These are used to determine if hard or soft boundaries should be used in grade interpolation. The contact plots are provided in Appendix C-3. Mineral Zone 3 has its only boundary with Mineral Zone 99; hence there is only one contact plot for Mineral Zone 3. All boundaries, except those between Mineral Zones 1 and 2, show sharp changes in grade profile going across the boundary. These should and were considered hard boundaries for interpolation. There are limited data near the contact for Mineral Zones 1 and 2, and the overall mean of Mineral Zone 1 is twice that of Mineral Zone 2. For this reason, the boundary was considered hard. If it had been considered soft, there would be danger of smearing grades in blocks near the boundary. A firm boundary could be considered for future models, but the number of blocks along the boundary affected would be small, and the effect limited.
17.7	Declustering
	To be effective, the block-size should be small enough so that all assays are assigned to some blocks. At Rock Creek, the nominal sample length is 1.5 m. But there are a significant number of assay intervals with different lengths. For example, for DDH data, there are 4,684 intervals with lengths of 1.50 to 1.53 m, 7,674 intervals with a length of 2 m, and 482 intervals with other lengths. For the RC data, there are 7,575 intervals with a length of 1.50 to 1.53 m, and 94 intervals with other lengths. Building a nearest neighbour model ignores this, i.e., assumes the assay intervals have the same length. If there is any relationship between sample interval length and grade, the nearest neighbour model and frequency distribution of grade may be biased.
	A secondary problem is the block size and data spacing. The block size is 10 × 5 × 5 m in the north-south, east-west and vertical directions. Given the assay intervals are nominally 1.5 m or 2.0 m in length, not all the assays will be assigned to blocks. The size of the Rock Creek deposit and lack of sub-blocking available in the MineSight® software, makes the use of smaller blocks impractical. For these reasons, the nearest-neighbour model method of declustering was rejected.
	Cell declustering is a very commonly used method. As normally implemented, the deposit is divided into large cells that contain multiple data points. The data captured in each cell are weighted by 1/n, where n is the number of data in a cell. The origin is often shifted, and the weights, over all cases are averaged to give a final weight to a

Project No. 151088	Page 17-5
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

datum. In a refined technique, the data are weighted by their length as well. This method works well in massive deposits that do not have boundaries, or have large distances between their extremes. When this is not the case, the effect of a cell that straddles a boundary, either lithological or topographic, can be disproportionate, as
•	the data density near the boundary is often sparse, and
•	the volume of the cell is much higher than the volume of mineralization within it.
AMEC personnel have seen the mean miscalculated by as much as 20% because of this; failure to length-weight can throw the mean off by 5%. As a result, AMEC developed a cell-declustering program that avoids these problems. The program has the following steps:
For each block (10 × 5 × 5 m), identify all assay intervals within a search distance. Determine the total weight of captured assay intervals:

There are m assay intervals captured within the search.

For each assay interval, identify all the blocks within the same search distance. Find the weight for each assay:

A final normalized weight F_aj is calculated:

The F_aj are the declustering weights and add to 1.

The program samples the declustered frequency distribution at even intervals and makes a file for use in plotting quantile-quantile plots (Q-Q).

Project No. 151088	Page 17-6
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Declustered frequency distributions are provided in Appendix B-1. These distributions are highly skewed, and tend to show a straight line on log-probability plots, indicating their fit to a simple lognormal distribution. The tails are thicker for RC distributions, giving rise to their much higher mean grades compared to core distributions.

Table 17-6 provides summary statistics for these distributions; their datasets are described in more detail below.

Table 17-6: Declustered Statistics (Au g/t) for Assay Distributions

	RC		Core		Trench
Dataset	Mean	CV	Mean	CV	Mean	CV
RC 501	1.53	2.67	0.75	2.66	-	-
RC 502, 503, 505, 506, 511	0.75	4.27	0.47	3.83	-	-
RC 512 and 516	0.78	3.04	0.58	3.22	-
Trench – RC	0.54	3.36	-	-	0.55	2.09
Trench – Core	-	-	0.31	4.13	0.55	2.09

	The RC and core distributions have similar CV; the trench data have much lower CVs; perhaps the support (volume) for these samples is greater.
17.8	Quantile-Quantile (Q-Q) Plots (General)
	A quantile is like a percentile, only the step between quantiles may be different from 0.01 in frequency. If two distributions are the same, their individual quantiles should be the same, and when the quantiles of one distribution are plotted against another, they should follow a line y=x. If the points fall off the line, the distributions are dissimilar. If distribution A quantiles are plotted using x coordinates, and distribution B quantiles are plotted using y coordinates, then A will be biased high with respect to B if the point falls below the line y=x. Conversely B will be biased high with respect to A if the point falls above the line y=x.
	It is possible to fit an equation or piecewise equations to the Q-Q plot data, which permits transformation of the A distribution into an A’ distribution which is “identical” to the B distribution. As a check, the B distribution can be plotted against A’ to see that the Q-Q data follow the line y=x. Figure 17-1 shows a Q-Q plot before adjustment of the RC data. Figure 17-2 shows a Q-Q plot after adjustment.

Project No. 151088	Page 17-7
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 17-1: Sample Q-Q Plot Before Adjustment

Figure 17-2: Sample Q-Q Plot After Adjustment

The Q-Q plots can be affected by outlier values at high quantiles; these are ignored in the fitting of equations.

Q-Q plots are provided in Appendix B-2. These are identified by Run No. at the base of the plot.

Project No. 151088	Page 17-8
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	Large adjustments were required to conform to the data distributions. It was recognized that there are sampling issues that can only be resolved by using a bulk sample or additional twin holes to referee the methods. Therefore, in this study, the adjustment equations were fitted by eye, and no attempt was made to precisely match means or deal with outliers.
17.9	Twin Site Q-Q Plots
	AMEC initially prepared a series of Q-Q plots for the twin hole data to provide some understanding as to how different drilling campaigns compared with one another under control situations (e.g. ‘A’ scored data). Figure 17-3 shows a Q-Q plot that compares AGC’s 2004 RC 516 program (x-axis) with their 2004 316 core program (y-axis).
	Figure 17-3: Twin Hole Q-Q Plot

The quantiles shown in Figure 17-3 show a slight to moderate high bias for the RC holes relative to the core holes. A more pronounced high bias is illustrated in Figure 17-4, which is a Q-Q plot that compares gold grade quantiles between AGC’s 2002 RC program (campaign 512) with AGC’s 2004 core program (campaign 316) for A+B+C scored holes.

Project No. 151088	Page 17-9
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

The quantiles for the 512 campaign (x-axis) are clearly biased relative to the 316 core hole campaign (y-axis). All of the twin site Q-Q plots are shown in Appendix B-1.

Figure 17-4: 316 vs. 512 Twin Hole Q-Q Plot

17.10	Q-Q Plot Adjustments
	Adjusted gold assay item AUAJ1 is calculated as a result of the Q-Q plot adjustment analysis and are stored in the MineSight® assay file (rcrk11.dat). AUAJ1 is an assay item where rotary RC drilling is adjusted down to a core datum. An upside adjustment of cores by 25% for campaigns 301-316 in Mineral Zones 1, 2, 3, and 99 is stored in AUAJ2. Many of the adjustment calculations are made in a piecewise continuous fashion, with grade intervals having linear relationships with differing slopes. The batch file, adjust.bat, performs these calculations with the associated run files (see Appendix E-1).
17.11	Comparison of RC and Core Data
	The majority of the sampling at Rock Creek has been obtained by DDH (core).

Project No. 151088	Page 17-10
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	RC methods can sometimes result in biased samples. For example, in areas of high water flow, gold can be concentrated via placering in the hole; weak zones above the sample interval will ravel, and if high-grade, will upgrade the sample. Alternatively, where gold is concentrated in fines, these may be lost in the sample collection process.
	Coring through friable, weathered intervals can cause a loss of gold in fine fractions.
	At Fort Knox, Alaska, AMEC made a detailed study of core and RC drilling; in the near- surface, dry, weathered environment, core samples were found to be biased low relative to a 40,000 t bulk sample. RC samples gave unbiased results. This zone extended to a depth of approximately 50 m. From 50 to 100 m, both core and RC methods gave similar results; below 100 m large water inflows were encountered, and RC samples were often upgraded by placering and contamination.
	At Rock Creek, core samples inside the Albion Shear gave similar results to RC samples, and no adjustment was required (see Runs 2-4, Appendix B-2). Outside the Albion Shear, core samples are low-biased with respect to RC samples. There is a limited dataset (approximately 908 samples) from trenches that are intermediate, i.e., lower than RC samples and higher than core samples in the vicinity. Photographs of the trenches show increased fine ravel; it is possible that fines could have been lost in the coring process. It is noteworthy that at Fort Knox, though the core recovery was approximately 80% near surface, the grade recovered by core samples was half that of a surrounding bulk sample.
17.12	Adjustment of RC to Core Datum
	It was suspected that the adjustment could be related to the drilling campaign. The 2004 summer DDH campaign 316 plus the 2003 campaign 315 were used as a reference for adjustment because it was felt that these would form the most reliable core dataset. Experimentation showed that the RC data could be divided into three groups:
	•	501 – The 1987 Placer Dome Drilling
	•	502, 503, 505, 506, 511 – Other holes drilled from 1988 to 1999
	•	512 + 516 – AGC 1999 and 2004

Project No. 151088	Page 17-11
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	•	For the 501 campaign, DDH is about 50% lower than RC. The adjustments for Runs 6 and 7 are progressively less severe, but still significant. The adjustment equations for RC samples located outside of the Albion Shear Zone are as follows:
	•	501 – The 1987 Placer Dome Drilling
		If < 1 g/t: Original Assay * 0.6; If >= 1 g/t: Original Assay * 0.3 + 0.3 g/t
	•	502, 503, 505, 506, 511 – Other holes drilled from 1988 to 1999
		If < 0.9 g/t: Original Assay * 0.75; If >= 0.9 g/t: Original Assay * 0.6 + 0.135 g/t
	•	512 + 516 – AGC 1999 and 2004
		If < 0.8 g/t: Original Assay * 0.8; If >= 0.8 g/t: Original Assay * 0.75 + 0.04 g/t.
	To review the sensitivity Q-Q plots related to the drilling campaigns mentioned above, refer to Appendix B-2 and B-3.
17.13	Observations
	Based on various comparisons (profile plots, basic descriptive statistics, and Q-Q plots) that were made for the twin site data the following observations were made:
	•	The 2003 and 2004 AGC core drilling campaigns (315 and 316 campaigns, respectively) have remarkably similar distributions (means, standard deviations and CVs). For twinned holes the mean grade of the 2003 core program is slightly higher than the mean of the 2004 program.
	•	Older RC campaigns show a clear high bias when compared to the 2003 and 2004 AGC core drilling programs.
	•	AGC’s 2004 RC data (campaign 516) shows a slight systematic high bias relative to the 2003 and 2004 core campaigns. This is shown by Q-Q plots and by the fact that the grade of the 516 campaign is about 12% higher than the 315+316 core campaigns for twinned holes.
	•	The 2004 RC campaign (516) has lower grades than the older RC campaigns (i.e. 502 through 511).
	•	The 2003 trench data fall between core (lower) and RC (higher) results.
	•	The 2004 trench data are significantly higher than the core and RC data. Therefore trench data were given very limited influence in resource estimation.

Project No. 151088	Page 17-12
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	Because core recovery is better than RC weight recovery, the base case for resource estimation will use the adjusted RC values outside the Albion Shear Zone – see equations in Section 17.12.
	AMEC reviewed the QA/QC data associated with the exploration and resource delineation drilling program at Rock Creek and found the data to be sufficiently accurate and precise for resource estimation. A brief summary of the procedures followed is given below.
17.14	Variograms for Assays
	Variograms were made using the 1 – correlogram method. This method produces a unit sill, with each lag normalized by dividing the covariance by the variance of the data. This method has been widely used for precious metals deposits for the last 15 years. It is robust to outliers and gives good results where data are limited.
	Variograms were calculated with azimuth and dip increments of 30°. Single structure models were fitted respecting the dip of the sheeted tension veins (dipping west at 75°). Variograms for assays are shown in Appendix D-1 and for composites in Appendix D-2. These are computed on the assay variable AURJ1, which is the RC adjusted to core datum grade variable. To damp noise, assays grading over 30 g/t or with lengths less than 1.4 m or greater than 6 m were not used.
	The variograms have high nugget effects and very short ranges. Table 17-7 describes the parameters for unit sill, exponential models. The ranges are “practical ranges”, i.e., they are distances at which the variogram reaches 95% of the sill.
	Table 17-7: Variogram Models for Assays (AURJ1 g/t)

Mineral Zone	Nugget Effect (C0)	C1	Range N-S	Range E-W Dip 75 W	Range E-W Dip 15 E
1	0.73	0.27	22.1	13.2	9.3
2	0.73	0.27	31.0	24.0	13.4
3	0.64	0.36	112.7	19.8	37.0
10	0.45	0.55	32.7	15.4	5.1
99	0.84	0.16	43.8	7.3	14.2

In particular, the nugget effects are relatively high for all zones except the Albion Shear Zone (Mineral Zone 10). This is typical of shear zone and vein-style precious metal mineralization. The models for all Mineral Zones show most continuity in the strike

Project No. 151088	Page 17-13
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	direction. The ratio of strike to down-dip ranges are similar for Mineral Zones 1, 2, and 10. Mineral Zones 3 and 99 show much more continuity in the strike than down-dip direction. According to AGC geologists, this may be explained by the fact that the veins are often brecciated and not oriented in a preferential direction except along the strike of the zone. The model ranges for Mineral Zone 99 are tenuous because of the extremely high nugget effect.
17.15	Variograms of 5 m Composites
	Variograms for drillhole composites are provided in Appendix D-2. The methodology for calculating composite variograms was the same as for assays (i.e. correlograms). To dampen noise from the correlograms, composites in excess of 10 g/t were not used. A minimum length of 2.5 m was required.
	The nugget effects for variogram models were fitted considering the nugget effect for assays, as shown in Table 17-8.
	Table 17-8: Derivation of Nugget Effects for Composite Variogram Models

Mineral Zone	Col 1	Col 2	Col 3	Col 4	Col 5	Col 6	Col 7	Col 8
	Co Assays (USVM)	CV Assays	Rel. Variance Assays	Rel. Co Assays	Calc’d Rel. Co Composites	CV Composites	Rel. Variance Composites	Co Composites (USVM)
1	0.73	2.446	5.983	4.368	1.572	1.531	2.344	0.671
2	0.73	3.431	11.772	8.593	3.094	2.215	4.906	0.631
3	0.64	4.937	24.374	15.599	5.616	3.428	11.751	0.478
10	0.45	2.941	8.649	3.892	1.401	1.951	3.806	0.368
99	0.84	4.060	16.484	13.846	4.985	2.452	6.012	0.829

•	Column 1 is the nugget effect for the assay variogram model
•	Column 2 is the CV for assays
•	Column 3 is the relative variance for assays = (Col 2)2
•	Column 4 is the relative C0 for assays = (Col 1)(Col 3)
•	Column 5 is the calculated relative C0 for composites = (Col 4)/[1.525/5]; support for assays = 1.525; support for composites = 5
•	Column 6 is the CV for composites
•	Column 7 is the relative variance for composites = (Col 6) 2
•	Column 8 is the nugget effect for the composite variogram model.
The remaining parameters of the composite variogram models were fitted using the experimental data. Variograms generated from composited data tend to dampen the nugget effect, allowing better revelation of spatial variability and estimation of ranges and anisotropy than raw uncomposited data.

Project No. 151088	Page 17-14
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 17-9 summarizes the fitted variogram models for each Mineral Zone. These are exponential models with practical ranges shown. Each practical range was divided by 3 prior to input into MineSight^® since this software does not use the practical range.

Table 17-9: Variogram Models for 5 m Composites

			Exponential Distances from			Distances from M624V1
	Nugget			SAGE
Mineral Zone	Effect	Sill (C1)	Range	Range	Range	Range	Range E-	Range
	(C0)	(C2)	N-S	E-W Dip	E-W Dip	N-S	W Dip	E-W Dip
				75 W	15E		75W	15 E
1	0.671	0.329	37.8	28.6	13.8	12.6	9.5	4.6
2	0.631	0.369	66.7	54.4	26.5	22.2	18.1	8.8
1+2	0.651	0.349	84.7	51.7	24.9	28.2	17.2	8.3
3	0.478	0.522	124.6	41.6	77.7	41.5	13.9	25.9
10	0.368	0.632	48.6	36.8	15.6	16.2	12.3	5.2
99	0.829	0.171	79.2	20.0	82.1	26.4	6.7	27.4
99 (0.4 g/t indicator)	0.500	0.500	26.5	13.8	12.2	8.8	4.6	4.1
99 (Low -grade <	0.650	C1=0.268	12.8	45.6	46.6	4.3	15.2	15.5
0.4 g/t)		C2=0.082	243.5	2191.6	86.7	81.2	730.5	28.9
99 (High-grade >	0.800	0.200	10.0	10.0	10.0	3.3	3.3	3.3
0.4 g/t)

17.16	Development of Target Coefficients of Variation for Resource Model
	AGC provided a target selective mining unit (SMU) of 10 × 10 × 5 m. This is the smallest volume that can be effectively segregated as ore or waste. Using the grade variogram models, it is possible to calculate the dispersion variances for SMUs. These can be used to develop target CVs for SMUs. If the CVs of resource block grades are similar, then the grade-tonnage curve of resource block grades will be representative of that which would apply during mining. Table 17-9 shows the calculations.
	These calculations take into account the support effect. Usually, there is also an information effect that further smoothes the grade-tonnage curve and has the impact of reducing the target CV. With such a large SMU, the impact of the information effect compared to the support effect is usually small. Conditional simulation should be considered in the future to provide insight into the information effect.
17.17	Gold Grade Interpolation Runs
	Six gold grade models were created, three of which had no metal removed and three had metal removed. The basic models were based on: (1) unadjusted grades where rejected RC and core intervals were not used (AUREJ), (2) RC adjusted to core datum grades (AUAJ1), and (3) core adjusted to RC datum grades (AUAJ2). The AUREJ

Project No. 151088	Page 17-15
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

model was used as the basis for calibrating metal removal as determined by the metal-at-risk study. The AUAJ1 model was estimated using RC assays that were adjusted relative to core data. This is the model that AMEC recommends should be used for pit optimization and economic studies. The AUAJ2 model represents an upside case where all core campaigns were factored upwards based on core-RC comparisons.

All three of the grade models were interpolated using the same parameters and techniques. Mineral Zones 1, 2, and 3 were interpolated using ordinary kriging methods. The Albion Shear Zone, (i.e. Mineral Zone 10) was estimated using inverse distance weighting methods (third power). The inverse distance method was selected because grade-tonnage curves from an ordinary kriged model could not be made to approximate a Herco model (it was too smooth). An indicator approach was used for the Mineral Zone 99, the default zone.

For each of the models, a three-pass interpolation method was used. Each zone was estimated with an initial pass that only required a minimum of one drillhole to be used. This pass was run to ensure that block grades were estimated around isolated holes. A second pass was run requiring at least two or more drillholes to be used. This run overwrote many of the blocks that had been estimated by the first pass. Both the first and second pass interpolation runs only used core and RC data. A third pass was then run that used trench and drillhole data provided the blocks were within a 15 x 15 x 5 m distance from trench data.

Hard boundaries were used for interpolating block grades for Mineral Zones 1, 2, 3, and 10. This means that strict zone matching between blocks and composites was imposed. Composites shorter than 2.5 m in length were not used. The grade estimate was weighted by the length of the composites.

The first two estimation passes used a 50 x 50 x 10 m search ellipse that was oriented north-south and dipped steeply to the west at 75 degrees. This tight ellipse was used in an attempt to orient the geometry of the mineralized zones in steep orientations that have been observed and modeled by AGC geologists.

An indicator cutoff of 0.40 g/t was selected for Mineral Zone 99 based on the proportions of metal above and below that cutoff grade. The indicator item INAU1 in the block model was kriged using 0’s and 1’s in the composite file that were set whether the composite was below or above a 0.40 g/t cutoff grade. Block grades were then estimated for the low and high grade populations using separate kriging runs. The final Mineral Zone 99 grade was calculated using the estimated indicator

Project No. 151088	Page 17-16
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

probability (INAU1) and the two gold grade fractions (AULO and AUHI) using the following formula:

Estimated Block Grade = (Indicator proportion)*(kriged high grade) + (1-indicator proportion)*(kriged low grade)

The number of composites and drillholes that were used to estimate each block were captured along with the distance to the closest drillhole composite that was used. In addition, a nearest neighbour grade was estimated for each model using the identical parameters that were used for the kriged or inverse distance runs.

Table 17-10 summarizes the six resource models that were constructed and shows what composite and model grade items were used.

Table 17-10: Grade Models

Model Number	Composite Grade Item	Model Grade Item	NN Model Grade Item	Description
1	AUREJ	AUKR1	AUNN1	Unadjusted grades - no metal removed
2	AUREJ	AUKR2	AUNN2	Unadjusted grades - metal removed
3	AUAJI	AUKR3	AUNN3	Adjusted grades - no metal removed
4	AUAJI	AUKR4	AUNN4	Adjusted grades - metal removed
5	AUAJ2	AUKR5	AUNN5	Upside grades - no metal removed
6	AUAJ2	AUKR6	AUNN6	Upside grades - metal removed

	A single batch file (make_models.bat) was used to estimate grades for all six models. This batch file and all pertinent MineSight® runfiles are contained in Appendix E-2.
17.18	Metal-at-Risk
	Precious metals deposits have skewed grade distributions. Skewed grade distributions have the property that a small proportion of samples can represent a disproportionately large amount of metal. The limited number of these samples can introduce significant uncertainty into a resource estimate. It is a common practice to cut the grades of very high-grade samples, restrict their projection distance, or to adjust resource models to mitigate downside risk.
	In many precious metals deposits, Rock Creek included, the highest-grade samples are scattered and discontinuous at the exploration drill-hole spacing. The number of high-grade samples intersected can vary according to the positioning of the drillholes,

Project No. 151088	Page 17-17
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

and it is impossible to know in advance which positions would give the most accurate estimate of the amount of high-grade metal actually present. The uncertainty related to the amount of high-grade metal can be evaluated using a Monte Carlo simulation technique developed by Mineral Resources Development/AMEC that has been applied over a 14-year period. This method essentially re-drills the deposit 1,000 times and notes the variation in the amount of high-grade metal present in annual or global production increments. The 20th percentile of the simulated metal contents is added to the metal content represented by the remaining samples to give a risk-adjusted metal content. The difference between total metal content and risk-adjusted metal content is termed metal-at-risk. Theoretically, in four periods out of five, the mine should do better than the estimate; however, there is additional and largely unquantifiable uncertainty related to the representivity of the sample-grade frequency distribution input to the simulation.

The appropriate time-period for a feasibility study stage of a project is annual, as Indicated Resources will be used to prepare annual production schedules. For Rock Creek, AGC has specified that the operation would produce 7,000 short tons per day or 2.5 million short tons per year.

The method has advantages over other top-cutting methods in that it takes into account 1) the data density, and 2) the volumes of increments used for production scheduling. As the data density is increased, the amount of metal-at-risk declines; longer production increments will have less risk than shorter ones.

The simulations were run using declustered grade distributions for AUCO assays (i.e., with RC samples adjusted to core datum).

Table 17-11 shows the results of the simulations.

Table 17-11: Metal-at-Risk by Mineral Zone (Indicated and Inferred)

				Expected High-
Mineral	Tons/	No. Currently	High-Grade	Grade		Metal	Metal-at-Risk
Zone	Assay	Available	Threshold	Assays in Period		Represented	(%)
		Assays Mined	(g/t Au)
		in Year		(No.)	(%)
1	3683	684	13	6	0.9	14.3	5.3
2	3683	684	9	5	0.7	13.8	7.6
3	3683	684	25	4	0.6	22.4	10.8
10	3683	684	25	5	0.7	18.2	7.3
99	8286	304	7	3	1.1	16.1	8.4

Project No. 151088	Page 17-18
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	The metal-at-risk is typical of scoping/prefeasibility stage projects. Typically with infill drilling (feasibility stage), the metal-at-risk for Indicated Resources is approximately 5%. This is approximately achieved for Mineral Zone 1. For the other zones the data density is sparse (more than 8,000 st/assay) and/or the distributions are more highly skewed than is typical (CV of assays higher than 2.0). Mineral Zones 3 (Walsh) and 99 are sparsely drilled. Mineral Zones 2 and 10 are relatively well drilled, but the CVs are 2.2 and 2.0, respectively.
	As explained in Section 17.17, metal-at-risk is removed in the modeling process by restricting the use of very-high grade composites in the estimation process.
17.19	Metal-at-Risk Adjustments
	Gold metal considered to be at risk was removed using the outlier restriction method during grade interpolation. No raw assay or composite data were capped, but rather the projected distance for certain data was restricted. Table 17-12 summarizes how target percentages of gold metal were removed from the base case model (AUKR4). The tonnes, grade, and contained gold ounces for the AUKR3 model are shown at the left side of Table 17-4 for each Mineral Zone. Using the outlier restriction parameters shown in the table, the tonnes, grade, and contained gold ounces for the AUKR4 model are compared to the AUKR3 model. As can be seen the actual metal that was removed for each mineral zone closely approximates the percentages targeted by AMEC’s Monte Carlo simulation method.

Table 17-12: Metal-at-Risk Removal

MINZN	AUKR3 (no metal removed)			AUKR4 (metal removed)			Outlier Parameters		%	% Metal
									Target	Removed
									Removal
	run708.m03-mod03.rpt			run708.m04-mod04.rpt
	kTonnes	Au	kOzs	kTonnes	Au	kOzs	Grade	Maxd
1	10,696	0.8362	288	10,696	0.7946	273	7.00	12.50	5.3%	5.0%
2	12,957	0.3964	165	12,957	0.3670	153	6.00	12.50	7.6%	7.4%
3	3,082	0.4722	47	3,082	0.3956	39	7.00	15.00	16.8%	16.2%
10	6,518	1.1794	247	6,518	1.0924	229	10.75	12.75	7.3%	7.4%
99	932,991	0.0176	528	932,991	0.0162	486	0.70 factor		8.4%	8.0%

Outlier restriction is a method of metal removal that permits composites above a nominated “outlier grade” to be seen at their full grade only by blocks within a search distance often much smaller than the regular search parameters. The outlier parameters were determined iteratively by selecting a grade/distance pair and determining the amount of metal removed relative to the target.

Project No. 151088	Page 17-19
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

17.20	Density
	Golder Associates performed density measurements on solid pieces of core data. The average density as determined by Golder4 was 2.73 g/cm3. Approximately 8% of the deposit consists of gouge, shears or fault material. It is prudent to assume that the density would be 10% lower in these zones. Adjusting for these material types suggests that a density of 2.71 g/cm3 maybe more appropriate.
	AMEC obtained the raw density data from AGC and have summarized the densities by rock type and depth from the surface. These data are presented in Table 17-13 and Table 17-14.
	The average specific gravity (SG) shown in Table 17-13 and Table 17-14 differ slightly from Golder’s global estimate of 2.73. Based on AMEC’s experience with other deposits that contain similar host rocks, the final SG value of 2.71 g/cm3 is appropriate for the estimate of tonnes.
	Table 17-13: Density by Rock Type

Rock Type	No. Determinations	SG (g/cm3)
Albion QTZ	8	2.70
CQMS	94	2.79
CS	44	2.76
GQMS	60	2.81
GQS	1	2.70
GS	5	2.71
MBL	28	2.74
QAB	2	2.90
QGS	19	2.70
QMS	83	2.80
QTZ	1	2.63
V3	16	2.69
V5	8	2.73
Total	369	2.77

______________________________
4 Geotechnical Investigation Proposed Rock Creek Mine Development Near Nome, AK, Golder and Associates, August 2004, Appendix E-4

Project No. 151088	Page 17-20
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 17-14: Density by Depth from Surface

Depth (m)	No. Determinations	SG (g/cm3)
0 to 10	27	2.77
0 to 20	33	2.79
0 to 30	46	2.78
0 to 40	37	2.78
0 to 50	40	2.75
0 to 60	36	2.79
0 to 70	33	2.77
0 to 80	26	2.78
0 to 90	28	2.76
0 to 100	22	2.80
0 to 110	12	2.77
0 to 120	8	2.78
0 to 130	6	2.76
0 to 140	6	2.73
0 to 150	4	2.63
0 to 160	1	2.78
0 to 170	2	2.77
0 to 180	1	2.84
0 to 200	1	2.75
Total	369	2.77

17.21	Resource Classification
	Mineral resources were classified based on a combination of distance to data and the number of drillholes that were used to estimate each block. The model item containing the resource classification code is called CLASS. At this time, no resources are considered to be Measured because of the difference of assay grades between RC samples and core samples. Blocks located within Mineral Zones 1, 2, 3, and 10 were classified as Indicated (CLASS = 2) or Inferred (CLASS = 3) if certain criteria were met. Blocks located in Mineral Zone 99 were classified as Inferred. Additional drilling and geologic modeling may permit portions of Mineral Zone 99 to be upgraded to Indicated resources.
	Table 17-15 summarizes the criteria used for classifying resources for each mineral zone.

Project No. 151088	Page 17-21
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 17-15: Indicated Resource Criteria

Mineral Zone	No Drillholes	Maximum Allowable	Distance to Closest
		Composite Distance (m)	Composite (m)
1,3,10	3 or more	45	0 to 23
	2 or more	33	0 to 15
	1	10	0 to 10
2	3 or more	35	0 to 18
	2 or more	25	0 to 13
	1	10	0 to 10

	All estimated blocks that were not classified as Indicated resources were classified as Inferred resources. A batch file (CLASS.BAT) was used to classify the resources, this file and all other pertinent MineSight® runfiles are included in Appendix E-3.
17.22	Model Validation
	The resource models were validated in several ways. Nearest neighbour models were built using 5 m composites for the base-case model using AUAJ1. Histograms were constructed for the kriged and nearest neighbour models (Appendix F-1 and F-2). The means for the kriged models and nearest neighbour models are provided in Table 17- 16 for Indicated Resources only. The comparisons are good for Mineral Zones 1, 2, and 10. For Mineral Zone 3 the Indicated Resource comparison is fair. This is understandable since the number of composites for Mineral Zone 3 is limited, and the Indicated tonnage is only 718 kt. For Indicated + Inferred the comparison is better.
	Table 17-16 shows the CVs for AUKR4 and the target CVs. There is a good comparison for Mineral Zone 1, which is important because it contains both ore and waste. For Mineral Zones 2, and 10, the CVs for kriged block grades are below target. It is very likely that tonnage will be overestimated and grade underestimated. In Mineral Zone 10, nearly the whole zone will be above cutoff (0.5 g/t), and there will be very little additional dilution incorporated into the estimate. Mineral Zones 2 and 3 are either minor in terms of ore tonnage or mainly Inferred. The data are so wide-spaced in the Inferred zones that it is usually impossible to match the CV target.
	The core-rotary sampling discrepancy is over-riding compared to these observations.

Project No. 151088	Page 17-22
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Table 17-16: Kriged Validation Against Nearest Neighbour Models (Indicated Only)

	Kriged Model Variable AUKR4			Nearest Neighbour
Mineral				Model Variable AUNN4
Zone	Mean	CV	Target CV	Mean	CV
			From Tab 7-3
1	0.841	0.727	0.821	0.805	1.599
2	0.443	0.983	1.185	0.437	2.177
3	0.586	1.609	1.787	0.565	3.376
10	1.346	1.274	1.628	1.413	2.123

Grade profiles were calculated for the nearest neighbour and kriged results summarized in Table 17-16 by averaging the blocks occurring in slices oriented on the coordinate axes (Appendix F-3).

In general, the comparisons show reasonable agreement and little recognizable local bias. The kriged estimate averages tend to smooth through the peaks and troughs of the nearest neighbour estimate. This is to be expected, as kriged estimates are weighted averages of data within the kriging neighbourhood. In some plots, divergence between the profiles is evidenced; however, these occurrences are generally coupled with low block counts that discourage any meaningful inference. The exception is Mineral Zone 99 where the comparisons are somewhat less in agreement although still acceptable, since the only resource claimed here is Inferred.

Figure 17-5 through Figure 17-7 show typical examples of these comparisons for the Indicated blocks for Mineral Zone 2.

Project No. 151088	Page 17-23
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 17-5: Example Profile – Gold Grade vs. Elevation

Project No. 151088	Page 17-24
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 17-6: Example Profile – Gold Grade vs. Northing

Project No. 151088	Page 17-25
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Figure 17-7: Example Profile – Gold Grade vs. Easting

Project No. 151088	Page 17-26
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Cross sections and plan maps were plotted and examined to compare the grade estimate to composited sample grades (Appendix F-5). The examination revealed that the kriging plan appropriately constrained high-grade samples and that the hard contacts between Mineral Zones were honoured. At Section 900 North, higher grade composite grades appear to have been overextended by a drillhole at the margin of estimation in Mineral Zone 99. The next model should employ the use of quadrant searching to control the amount of extrapolation permitted in the unconstrained Mineral Zone 99 domain.
An independent check on the smoothing in the estimates was made using the Discrete Gaussian or Hermitian polynomial change-of-support method. The distribution of hypothetical block grades derived by this method is compared to the estimated model grade distribution by means of grade-tonnage curves. The grade-tonnage curves allow comparison of the two grade distributions in a format familiar to mining. If the estimation procedure has adequately predicted grades for the selected block size, then the grade-tonnage curves should match closely. If the curves diverge significantly, then there is a problem with the estimated resource.
This method uses the “declustered” distribution of composite grades from a nearest- neighbour or polygonal model to predict the distribution of grades in blocks. In this case, the blocks used in the model are 10 x 10 x 5 m. The unadjusted polygonal model assumes much more selectivity for ore and waste than is actually possible in mining practice, since many sample-sized volumes are averaged together within a block. This means that part of the sample-sized volumes in the block may be ore (above the mining cutoff) and part may be waste. Hence, the distribution of the grade of the blocks is not likely to resemble the distribution of grades from composite samples derived from the polygonal estimate. The method assumes that the distribution of the blocks will become more symmetric as the variance of the block distribution is reduced, i.e., as the mining blocks become bigger.
The histogram for the blocks is derived from two calculations:
•	the block-to-block variance (sometimes referred to in statistics as the between- block variance), which is calculated by subtracting the average value of the variogram within a block from the variance for composite samples (the sill of the variogram)
•	the frequency distribution for the composite grades transformed by means of hermite polynomials (Herco: hermite correction) into a less skewed distribution with the same mean as the declustered grade distribution and with the block-to-block variance of the grades.

Project No. 151088	Page 17-27
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

	The distribution of hypothetical block grades derived by the Herco method is then compared to the estimated grade distribution to be validated by means of grade- tonnage curves.
	The distribution of calculated 10 x 10 x 5 m block grades for gold in the domains are shown with red lines on the grade-tonnage curves in Appendix F-4. This is the distribution of grades based on SMU blocks obtained from the change-of-support models. The black lines in the figures show the grade-tonnage distribution obtained from the block estimates. The grade-tonnage predictions produced for the models show that grade and tonnage estimates are validated by the change-of-support calculations over the likely range of mining grade cutoff values (0.4 to 1.0 g/t). At higher cutoff grades in Mineral Zone 3, the kriged model gives lower grades and higher tonnages.
	AMEC recommends that AGC examine the default Mineral Zone 99 to determine if areas within it have been adequately drilled to warrant geologic interpretation and construction of new mineral zones.
17.23	Resource Statements
	The resource estimate was updated by AMEC from October to December 2004. Adjustment formulas were reviewed in light of the extensive core and rotary twin drilling program completed in the summer of 2004. Analysis of the twins largely confirmed previous conclusions except that no adjustment of RC appears necessary within the Albion Shear. Rotary data were adjusted to core datum outside of the Albion Shear Zone. Sensitivity cases were developed using unadjusted RC data.
	Table 17-17 provides a summary of mineral inventories at various cutoff grades for the different cases. Appendix G provides detailed statements for each case. Model definitions are provided at the bottom of Table 17-17.
	There is considerable difference between the models. Model 4 has been constructed with RC data adjusted to the core datum and metal-at-risk removed. This case is recommended for mine planning, based on our current knowledge.
	At the request of NovaGold, Mike Lechner of Resource Modeling Inc. (RMI) generated a Lerchs-Grossmann pit shell using a US$500/oz gold price and other parameters taken from Norwest (2005). Mineral Resources within this pit are declared using Model 4:

Project No. 151088	Page 17-28
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

•	9.595 million tonnes (Indicated) at a grade of 1.31 g/t containing 404,000 oz Au.
•	1.432 miliion tonnes (Inferred) at a grade of 0.96 g/t containing 44,000 oz Au.
The mineral resources of the Rock Creek deposit were classified using logic consistent with the CIM definitions incorporated in NI 43–101. The mineralization of the project satisfies sufficient criteria to be classified into the Indicated mineral resource category.
Table 17-17: Mineral Inventories (Model 4 is the Base Case Recommended for Planning)

Au	Resource	Indicated			Inferred			Contained Metal Ratio vs Base
Cutoff	Model							Case Model (i.e. Model 4)
(g/t)
		Tonnes	AU (g/t)	Au Ozs	Tonnes	Au (g/t)	Au Ozs	Indicated	Inferred
		(‘000)		(‘000)	(‘000)		(‘000)
	Model 1	25,638	0.88	725	50,179	0.31	500	1.18	1.12
	Model 2	25,633	0.82	676	49,870	0.29	465	1.10	1.04
0.10	Model 3	25,587	0.81	666	50,122	0.31	500	1.08	1.12
	Model 4	25,581	0.75	617	49,813	0.28	448	1.00	1.00
	Model 5	25,743	0.96	795	50,284	0.31	501	1.29	1.12
	Model 6	25,740	0.89	737	49,975	0.29	466	1.19	1.04
	Model 1	23,767	0.94	718	27,770	0.44	393	1.19	1.15
	Model 2	23,726	0.87	664	26,017	0.41	343	1.10	1.01
0.20	Model 3	23,277	0.87	651	27,659	0.44	391	1.08	1.15
	Model 4	23,234	0.81	605	25,906	0.41	341	1.00	1.00
	Model 5	24,398	1.00	784	28,046	0.45	406	1.30	1.19
	Model 6	24,361	0.93	728	26,293	0.42	355	1.20	1.04
	Model 1	20,726	1.04	693	15,585	0.60	301	1.21	1.24
	Model 2	20,642	0.96	637	13,458	0.57	247	1.11	1.02
0.30	Model 3	19,660	0.99	626	15,387	0.60	297	1.09	1.22
	Model 4	19,562	0.91	572	13,261	0.57	243	1.00	1.00
	Model 5	21,780	1.09	763	16,028	0.60	309	1.33	1.27
	Model 6	21,712	1.01	705	13,901	0.57	255	1.23	1.05
	Model 1	17,533	1.17	660	9,796	0.75	236	1.23	1.29
	Model 2	17,366	1.08	603	8,098	0.72	187	1.12	1.02
0.40	Model 3	16,404	1.11	585	9,598	0.75	231	1.09	1.26
	Model 4	16,211	1.03	537	7,900	0.72	183	1.00	1.00
	Model 5	18,875	1.21	734	10,135	0.75	244	1.37	1.33
	Model 6	18,755	1.12	675	8,437	0.72	195	1.26	1.07
	Model 1	14,909	1.29	618	6,837	0.89	196	1.25	1.35
	Model 2	14,667	1.19	561	5,542	0.86	153	1.13	1.06
0.50	Model 3	13,648	1.25	548	6,602	0.89	189	1.11	1.30
	Model 4	13,385	1.15	495	5,307	0.85	145	1.00	1.00
	Model 5	16,281	1.33	696	7,049	0.89	202	1.41	1.39
	Model 6	16,084	1.23	636	5,758	0.86	159	1.28	1.10
	Model 1	12,656	1.43	582	4,571	1.03	157	1.28	1.37
	Model 2	12,341	1.31	520	3,727	1.01	121	1.15	1.05
0.60	Model 3	11,369	1.39	508	4,608	1.03	153	1.12	1.33
	Model 4	11,040	1.28	454	3,583	1.00	115	1.00	1.00
	Model 5	14,052	1.45	655	4,938	1.03	164	1.44	1.43
	Model 6	13,783	1.34	594	3,918	1.00	126	1.31	1.10
0.70	Model 1	10,623	1.57	536	3,677	1.15	136	1.29	1.40

Project No. 151088	Page 17-29
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Au	Resource	Indicated			Inferred			Contained Metal Ratio vs Base
Cutoff	Model							Case Model (i.e. Model 4)
(g/t)
		Tonnes	AU (g/t)	Au Ozs	Tonnes	Au (g/t)	Au Ozs	Indicated	Inferred
		(‘000)		(‘000)	(‘000)		(‘000)
	Model 2	10,244	1.45	478	2,771	1.13	101	1.15	1.04
	Model 3	9,464	1.53	466	3,602	1.14	132	1.13	1.36
	Model 4	9,072	1.42	414	2,695	1.12	97	1.00	1.00
	Model 5	12,017	1.59	614	3,809	1.15	141	1.48	1.45
	Model 6	11,688	1.46	549	2,904	1.13	106	1.33	1.09
	Model 1	8,988	1.72	497	2,724	1.29	113	1.32	1.43
	Model 2	8,561	1.59	438	1,990	1.28	82	1.16	1.04
0.80	Model 3	8,012	1.68	433	2,674	1.27	109	1.15	1.38
	Model 4	7,565	1.55	377	1,940	1.26	79	1.00	1.00
	Model 5	10,267	1.73	571	2,860	1.28	118	1.51	1.49
	Model 6	9,896	1.59	506	2,126	1.27	87	1.34	1.10
0.90	Model 1	7,657	1.88	463	2,127	1.41	96	1.35	1.45
	Model 2	7,182	1.73	399	1,519	1.41	69	1.16	1.05
	Model 3	6,853	1.82	401	2,089	1.40	94	1.17	1.42
	Model 4	6,365	1.68	344	1,481	1.39	66	1.00	1.00
	Model 5	8,771	1.88	530	2,209	1.41	100	1.54	1.52
	Model 6	8,355	1.73	465	1,601	1.40	72	1.35	1.09
1.00	Model 1	6,615	2.02	430	1,743	1.51	85	1.38	1.47
	Model 2	6,093	1.87	366	1,241	1.52	61	1.17	1.05
	Model 3	5,865	1.96	370	1,712	1.49	82	1.19	1.41
	Model 4	5,328	1.82	312	1,211	1.49	58	1.00	1.00
	Model 5	7,621	2.02	495	1,787	1.52	87	1.59	1.50
	Model 6	7,160	1.86	428	1,288	1.52	63	1.37	1.09

Model 1 – Unadjusted Grades – no metal-at-risk removed
Model 2 – Unadjusted Grades, metal-at-risk removed
Model 3 – RC adjusted to core, no Metal-at-risk removed
Model 4 – RC adjusted to core, metal-at-risk removed (Base Case) = reportable Mineral Resource
Model 5 – Core increased 25 %, no RC adjustment, no metal-at-risk removed
Model 6 – Core increased 25 %, no RC adjustment, metal-at-risk removed

17.24	Reconciliation to Previous Models
	AGC requested AMEC complete a reconciliation of resources between the AGC 2003 PACK model and two subsequent AMEC resource models. AMEC’s reconciliation is summarized below and the details are provided in Appendix I.
	The 2003 scoping study pit contains nearly three times as much total material as the 2004 LG pit (56.7 million tonnes vs. 18.6 million tonnes). Similarly, this pit contains roughly 40% more material above a 0.62 g/t cutoff grade than the other pits. The 2003 PACK model contains about 400,000 more contained gold ounces than the other AMEC models within the scoping study pit. Approximately half of the difference (200,000 oz) is due to the fact that the 2003 scoping study pit was optimized using all estimated blocks from the 2003 PACK model. Block grades were estimated up to

Project No. 151088	Page 17-30
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

100 m from drillhole data in that model. The maximum distance that drillhole grades were projected in the AMEC models was 50 m.
Differences in estimation methods account for the remaining 200,000 oz difference between the 2003 PACK model and the various AMEC models. In addition to estimation differences between the various models, the November 2004 AMEC model was estimated with significantly more drillhole data than were available for the 2003 PACK model. The additional drilling data consisted of core and RC twin holes and local infill data.
The difference in contained ounces between the two models, attributable to differing estimation methods, is difficult to quantify by individual categories. The principal differences include:
•	RC assays were factored downward by AMEC for the November 2004 model while no factoring of assays was done for the 2003 PACK model
•	Closer-spaced drilling data were available for the 2004 model
•	There are a number of cases where excessive grade projection occurred from high-grade intercepts at the bottom of drillholes in the 2003 PACK model
•	Localized grade smearing in the 2003 PACK model
•	The way in which metal-at-risk was handled for each model.
The decision by AMEC to factor RC assay data is discussed in the June 2004 report and in Section 17.12 of this report. There are several examples where 2003 PACK block grades could not be supported based on newly acquired infill drilling data (e.g. Section 480N). Similarly, in the 2003 PACK model, it was observed that the last drillhole composite in at least five drillholes generated a large volume of blocks in excess of 2 g/t gold. In these cases, the high-grade composites were only loosely constrained by a grade probability contour (0.37 probability of the block being in excess of 0.25 g/t). In the 2003 PACK model, metal-at-risk was removed by capping gold composites at 15, and 17 g/t for the tension vein regime (Mineral Zones 1 and 2) and the Albion Shear zone material, respectively. Metal-at-risk was removed by the outlier restriction method in the November 2004 AMEC model.

Project No. 151088	Page 17-31
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

18.0	OTHER RELEVANT DATA AND INFORMATION
	No other data or information is relevant for the review of the Rock Creek project.

Project No. 151088	Page 18-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

19.0	INTERPRETATIONS AND CONCLUSIONS
	In 2004, the last time AMEC Consultants visited the Rock Creek Property, the land title agreement between AGC and the aboriginal land owners had not been fully settled. In 2006, AGC has provided documentation to AMEC that the land title agreement is in good standing based on the report issued by the attorneys represented by Guess and Rudd, effectively dated August 11, 2006. Based on this knowledge, AMEC believes that any resource statement produced within the data reviewed in this report is appropriate for public disclosure.
	AMEC has reviewed the interpreted geology and the mineralization at the Rock Creek deposit. AMEC believes that the geology and mineralized zones have been adequately mapped and interpreted to complete a mineral resource.
	AMEC has reviewed the drilling and sampling methods used on the Rock Creek Property and believes that these methods and procedures are reasonably accurate and conform to industry-standard practices for the development of a resource declaration.
	AMEC has also reviewed the data verification along with the sample preparation and analyses procedures for assaying facilities used by AGC. AMEC believes that the results of the review are adequate for the development of a resource declaration, although there are some discrepancies that have been noted. Refer to the recommendations for details.
	AMEC concludes that the resource statement, based on adjustment of RC sample grades to the core “datum” and with metal-at-risk removed, is the most appropriate based on the current information. This resource statement incorporates the following features:
	•	Conservative resource reporting is recommended until production data reconciliation answers the bias discrepancy between the RC and core assay data.
	•	Due to the qualitative nature of core samples versus RC samples and their noted discrepancies, AMEC commends AGC’s implementation of RC assays adjusted to core assays for the resource estimation.
	•	The reduced significance of pre-1990 Placer Dome and Tenneco data with new RC and core drilling in the proposed pit area.
	•	The restricted use of trench data (blocks within 15 m of trench assays) in the resource estimation.

Project No. 151088	Page 19-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

•	The rejection of RC samples with less than 30% recovery and those with greater than 120% recovery used in the resource estimation.
•	The rejection of core samples with less than 60% recovery used in the resource estimation.
•	Currently, there is only an indicated resource applicable to the Rock Creek Deposit.
Removing the metal-at risk has created a reasonably conservative resource statement, effectively removing between 5% and 11% of the total resource based on not removing any metal-at-risk.
It is the opinion of AMEC that the review of the work performed has met the objective for AGC to declare a Mineral Resource for the Rock Creek Property. NovaGold should complete its feasibility studies to enable conversion of Mineral Resources to Mineral Reserves.

Project No. 151088	Page 19-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

20.0	ADDITIONAL REQUIREMENTS FOR TECHNICAL REPORTS ON DEVELOPMENT PROPERTIES AND PRODUCTION PROPERTIES
	There are no relevant additional requirements.

Project No. 151088	Page 20-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

21.0	RECOMMENDATIONS
	Following the audit of the AGC mineral resource estimate, AMEC recommends the following:
	•	AGC perform, in due course, the remaining comments and requirements made by Guess and Rudd to be assured that the land titles are settled and clearly defined. AGC has stated their intentions to act on these comments and requirements in a letter to AMEC dated August 14, 2006.
	•	Compare the results of the 84 re-surveyed drillhole coordinates (from October, 2004) with the original surveyed hole coordinates to make sure that the holes are within acceptable tolerances. This may have already been completed, but AMEC did not have access to any of the resultant comparisons.
	•	Conduct sampling studies to determine the loss of gold during sample preparation that was determined by the check assay campaign, based on coarse reject samples. The result of 77 check assays determined that there was a nine percent high bias at the check lab (SGS Lakefield). Based on this information, AGS Chemex may be understating gold values.
	•	Fully investigate the noted discrepancies in the master database. The data quality from 2003-2004 has been deemed excellent with only one error found in the nine holes audited. Data from before 2003, has a higher error rate and therefore should be fully audited by AGC to correct discrepancies noted in section 14.2
	•	The QA/QC data should be compiled into organized groups of datasets. A summary document on all QA/QC programs should be prepared for future resource models and technical reports.
	•	AGC should investigate the default Mineral Zone 99 to determine portions that have been adequately drilled to support a revised geological interpretation and construction of new mineralized zones.
	•	The next resource model should include a quadrant search to control the amount of extrapolation in the unconstrained Mineral Zone 99. This will force the estimation runs to use a minimum number of declustered samples from more than one drillhole, thus avoiding grade smearing into areas of limited drilling.

Project No. 151088	Page 21-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

22.0	REFERENCES
	Avalon Development Corporation, 2002. Summary Report for the Rock Creek Gold Prospect, Seward Peninsula, Alaska.
	Bundtzen, T.K., Reger, R.D., Laird, G.M., Pinney, D.S., Clautice, K.H., Liss, S.A., and Cruse, G.R., 1994, Progress Report on the geology and mineral resources of the Nome mining district: Alaska Division of Geological and Geophysical Surveys, Public Data-File 94-39, 21 pages, 2 sheets, scale 1:63,360.
	Cox, D.P., and Singer, D.A., eds., 1986, Mineral deposit models: U.S. Geological Survey Bulletin 1693, 379 p.
	DGGS, 1994a, Colour shadow total field magnetics of the Nome Mining District: Alaska Div. Geological Geophys. Surveys, Rept. of Inv. 94-11, one colour map.
	DGGS, 1994b, 7200 Hz coplanar resistivity of the Nome Mining District: Alaska Div. Geological Geophys. Surveys, Rept. of Inv. 94-12, one colour map.
	DGGS, 1994c, 900 Hz coplanar resistivity of the Nome Mining District: Alaska Div. Geological Geophys. Surveys, Rept. of Inv. 94-13, one colour map.
	Flanigan, B., Freeman, C., Newberry, R., McCoy, D., and Hart, C., 2000, Exploration models for mid and Late Cretaceous intrusion-related gold deposits in Alaska and the Yukon Territory, Canada, in Cluer, J.K., Price, J.G., Struhsacker, E.M., Hardyman, R.F., and Morris, C.L., eds., Geology and Ore Deposits 2000: The Great Basin and Beyond: Geological Society of Nevada Symposium Proceedings, May 15-18, 2000, p. 591-614.
	Ford, R.C., 1993, Geology, geochemistry, and age of gold lodes at Bluff, and Mt. Distin, Seward Peninsula, Alaska: unpublished Ph.D. dissertation, Colourado School of Mines, Colourado, 302 p.
	Ford, R.C., and Snee, L.W., 1996: 40Ar/39Ar thermochronology of white mica from the Nome district, Alaska -- The first ages of lode sources to placer gold deposits in the Seward Peninsula: Economic Geology, v. 91, no. 1, p. 213-220.
	Gamble, B.M., Ashley, R.P., and Pickthorn, W.J., 1985, Preliminary study of lode gold deposits, Seward Peninsula, IN Bartsch-Winkler, Susan, ed., The United States Geological Survey in Alaska – Accomplishments during 1984: U.S. Geological Survey Circular 967, p. 27-29.

Project No. 151088	Page 22-1
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

Goldfarb, R.J., 1997, Metallogenic evolution of Alaska: Econ. Geol. Mono. 9, pp. 4-34.

Goldfarb, R. J., Miller, L. D., Leach, D. L., and Snee, L. W., 1997, Gold deposits in metamorphic rocks of Alaska, IN Goldfarb, R.J., and Miller, L.D., eds., Mineral deposits of Alaska: Economic Geology Monograph9, p. 151-190.

Hawley, C.C., and Hudson, T.L., 1996: Distribution of mineral occurrences in the Nome 1:250,000-scale quadrangle, Alaska, USGS OPEN-FILE REPORT 02-113

Hummel, C.L., 1962a, Preliminary geology of the Nome C-l Quadrangle, Seward Peninsula, Alaska: U.S. Geological Survey Map MF-247., 1 sheet, scale 1:63,360.

Hummel, C.L., 1962b, Preliminary geology of the Nome D-l Quadrangle, Seward Peninsula, Alaska: U.S. Geological Survey Map MF 248, 1 sheet, scale 1:63,360.

Krzewinski, T.G., Ross, T.E. and Nicholson, D., 2004: Rock Creek Gold Mine Near Nome, Alaska Geotechnical Investigation/Studies for Open Pit, Tailings Impoundment, Waste Dumps, and Plant Facilities: unpublished internal report on Golder website.

Nokleberg, W.J., West, T.D., Dawson, K.M., Shpikerman, V.I., Bundtzen, T.K., Parfenov, L.M., Monger, J.W.H., Ratkin, V.V., Baranov, B.V., Byalobzhesky, S.G., Diggles, M.F., Eremin, R.A., Fujita, Kazuya, Gordey, S.P., Gorodinskiy, M.E., Goryachev, N.A., Feeney, T.D., Frolov, Y.F., Grantz, Arthur, Khanchuck, A.I., Koch, R.D., Natalin, B.A., Natapov, L.M., Norton, I.O., Patton, W.W., Jr., Plafker, George, Pozdeev, A.I., Rozenblum, I.S., Scholl, D.W., Sokolov, S.D., Sosunov, G.M., Stone, D.B., Tabor, R.W., Tsukanov, N.V., and Vallier T.L., 1998: Summary terrane, mineral deposit, and metallogenic belt maps of the Russian Far East, Alaska, and the Canadian Cordillera: U.S. Geological Survey Open-File Report 98-136 (CD-ROM).

Norwest Corporation (2003), Preliminary Economic Study – Rock Creek Project, Nome,Alaska.

Norwest Corporation (2005), Rock Creek Project Updated Economic Review

NovaGold Ltd., 2006: website www.novagold.com

Plafker, G. and Berg, H.C., 1994, Overview of the geology and tectonic history of Alaska in Plafker, G. and Berg, H.C., editors, The Geology of Alaska: Geological Soc. Amer., Geology of North Amer., V. G-1, pp 989-1021.

Project No. 151088	Page 22-2
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

St. George, P., 2000, 1999 Rock Creek project drilling and metallurgical program: Internal Rept., NovaGold Resources Inc., 29 p.

Till, A.B., and Dumoulin, J.A., 1994, Geology of the Seward peninsula and St. Lawrence Island: in Plafker, George, and Berg H.C., editors, The Geology of Alaska: Geological Soc. Amer., Geology of North Amer., V. G-1, pp 141-152.

Project No. 151088	Page 22-3
11 September 2006

	Alaska Gold Company, a subsidiary of
	NovaGold Resources, Inc.
	Rock Creek Project
	Resource Model 2004

23.0	DATE AND SIGNATURE PAGE
	The undersigned prepared this Technical Report, titled Technical Report of the Rock Creek Property, Nome Alaska, USA dated 30 of August 2006, in support of the public disclosure of technical aspects of the Rock Creek Project as of the 30 of August 2006. The format and content of the report are intended to conform to Form 43-101F1 of National Instrument 43–101 (NI 43–101) of the Canadian Securities Administrators.
	Signed
	Dr. Harry M. Parker                                          30th August 2006.

Project No. 151088	Page 23-1
11 September 2006