o

o o o

Strong Support from Government & Society Significant Milestones to First Gold Production Mineral Resources & Reserves Statement Technical Report

The Martabe Project, a world-class gold and silver mine, was first discovered 10 years ago in the Province of North Sumatra, Indonesia. Since then, the resource base of the Martabe Gold and Silver Project has continued to grow and is currently at 6.5 million ounces of gold and 66 million ounces of silver. The Martabe Project lies in a substantial area of a highly mineralised district with potential for further gold and gold-copper deposits. Currently, under the existing Contract of Work G-Resources has complete access to this area of 1,639 square kilometres, which is valid for a period of up to 50 years. The Martabe Project is conveniently located with excellent infrastructure, including the Trans-Sumatran highway, the nearby port town of Sibolga, a 230 MW coal fired power station and high-voltage grid, water resources, and existing communications and local business infrastructure.

Highlights The Martabe Project is a gold-silver development project located in North Sumatra Province, Indonesia:

3.03 million ounces of gold reserves and 33.63 million ounces of silver reserves 6.5 million ounces of gold resources1 and 66 million ounces of silver resources1 Part of a large-scale mineralised district which has the potential to host further gold and base metal deposits

Resources are inclusive of reserves

Project started construction in July 2008 and is expected to be complete in 2011

In mid 2010, anticipated total development capital expenditures of US$440 million, of which around US$150 million has already been spent

The Construction Permit for the project was granted by the Indonesian Authorities in April 2008.

The 1,639 km2 tenement, highlighting the Martabe Project area.

The Martabe Project benefits from a large and high quality Mineral Resource base, with Ore Reserves of 3.03Moz of gold and 33.63Mozof silver contained within the Purnama/Pit 1, Barani, and Ramba Joring deposits and Mineral Resources of 6.49Mozof gold and 66.2Moz of silver from the Purnama, Barani, Ramba Joring and Uluala Hulu deposits, as classified under the Australian JORC Code. Based on the current Ore Reserves, the life of mine is expected to be ten years with annual production for the early years of mine life being on average 250,000oz of gold and 2-3Moz of silver. We believe there is significant potential to extend the life of the mine and Mineral Resource base, given the potential for additional Mineral Resources already identified at adjacent deposits. Moreover, the low strip ratio of approximately 1.3 tonnes of waste material to each tonne of ore, together with the good ore mineralogy of the deposits, allows for a high metal recovery ratio, conventional open pit mining methods and low overall cash cost of production. Life of mine cash costs are estimated at US$242 per oz after silver credits, putting this operation in the lowest quartile of costs for global gold mines. Large Reserve and Resource Base

Reserves defined for Purnama, Barani and Ramba Joring: 3.03 Moz Au and 33.63 Moz Ag Resources defined for Purnama, Barani, Ramba Joring and Uluala Hulu prospects: 6.49 Moz Au and 66.2 Moz Ag

Martabe Mineral Resource estimates, 0.5 g/t Au cut-off grade. Deposit 0.5 g/t Au cut-off grade Pit 1 Category Tonnes Gold Grade Silver Grade (Mt) 4.36 36.44 41.24 82.04 10.30 6.60 16.90 33.71 4.64 38.35 0.77 0.49 1.26 138.60 (g/t Au) 3.1 2.0 1.3 1.7 1.3 1.1 1.2 1.0 0.8 1.0 2.3 1.5 2.0 1.5 (g/t Ag) 53 24 17 22 3.5 2.4 3.2 4.1 3.7 4.1 31 12 24 15 Contained Metal Silver Gold (Moz) (Moz) 0.43 7.46 2.36 28.46 1.74 22.54 4.53 58.46 0.42 1.14 0.24 0.63 0.66 1.77 1.11 4.47 0.12 0.55 1.23 5.02 0.06 0.77 0.02 0.18 0.08 0.95 6.49 66.20

Measured Indicated Inferred TOTAL Indicated Barani Inferred TOTAL Ramba Joring Indicated Inferred TOTAL Indicated Uluala Hulu Inferred TOTAL TOTAL

Martabe Ore Reserve Estimates Deposit Category Tonnes Gold Grade Silver Grade Contained MetalGold (Mt) (g/t Au) (g/t Ag) (koz) (koz) Proved 4.4 3.0 53.5 428 7,548 Pit 1 Reserves Probable 30.6 2.1 25.3 2,066 24,860 Reserves Probable 4.5 1.6 2.8 225 412 Barani Reserves Probable 6.0 1.6 4.2 315 810 Ramba Joring Reserves TOTAL ORE 45.5 2.1 23.0 3,034 33,630 RESERVES Significant figures do not imply precision. Figures are rounded according to JORC Code guidelines. Ore Reserves for Pit 1 and Barani are estimated using a gold price of US$820/oz and silver price of US$12.50/oz. Ore Reserves for Ramba Joring are estimated using a gold price of US$1,054/oz and silver price of US$17.50/oz.

Key areas for growth:

Purnama (Pit 1) Reserve Barani Resources Ramba Joring Resources Other Deposits where Resources are yet to be defined Near mine exploration targets

A geological setting fertile for large ore bodies COW covers significant mineral district

Multiple untested prospects in 1,639km COW area:

Numerous new gold targets Additional regional targets show porphyry potential

The Martabe Project has a low estimated cost structure, which compares favourably to competitors. Its low cost structure is largely due to the capacity and size of the project (4.5 Mtpa bulk style open-cut mining); the high grade of the ore (2.1 g/t gold average); good recovery (7080% for gold and 60-80% for silver); a very low strip ratio of approximately 1.3:1 (waste: ore); and straightforward mining processes (mine, crush, SAG/Ball mill, CIL operation). In addition, the project site benefits from the favourable location of facilities, which has the potential to reduce operating and transportation costs and provide access to a large base of skilled but low cost Indonesian workforce. As a result of these factors, the Martabe Projects operating life of mine cash cost is estimated to be approximately US$242/oz (net of by-product credits). In the early years of operation, the cash costs are expected to be lower. The Board views strict cost management and increases in productivity as fundamental aspects of operation and will continually seek to improve the efficiency and costs of the Martabe Project. Martabes operating life of mine cash cost is estimated at US$242/oz (net of by-product credits)

Very favourable cost position compared to its peers

2011 Estimated Gold C1 Cost

Straightforward, Low-Cost Mining and Processing

4.5 Mtpa bulk style open-cut mining; extremely low strip ratio of 1.3:1 Straightforward and standard process

Mine, crush, SAG/Ball mill, CIL operation

Anticipated total cash cost net of by-product credits: US$242/oz Operating Costs Martabes total production cost averages US$242/oz of gold over the life of the mine

Low costs attributable to: o Scale o Low Strip Ratio o Grade o Simple process flow-sheet o Good metallurgical recoveries o Excellent logistics and existing infrastructure o Existing Indonesian mining capability

Operating Costs Mine Plan Annual Average Cash Costs Mining Processing Maintenance Administration, refining and other Contingency Total Costs Royalties Silver Credit (at US$32/oz) Total Cash Cost Total Cash Cost / oz Project Schedule and Milestones Schedule G-Resources recommenced the project end 2009 Ausenco appointed EPCM contractor Engineering and procurement 95% complete Major processing plant items manufactured Grid power contract signed with PLN A number of major contracts awarded: PT Leighton Indonesia (earthworks, roads, tailings dam, concrete), PT Duta Graha Indah (dams, roads, concrete supply), PT Grama Bazita (infrastructure), PT Lintech (steelwork), PT Epiterma Mas (tanks)

US$M 40 52 10 20 11 133 2 -77 58 US $242 / oz

Milestones Mining overburden removal First ore mined to stockpile Award of steel erection, mechanical & piping installation contract Award of electrical and instrumentation installation contract Grinding SAG mill delivery to site Ore primary crusher delivery to site CIL processing tank erection complete TSF Phase I complete for receipt of tailing Wet commissioning start First gold Sustained bullion production Project Capital Expenditures Capital Expenditures Total Project Capital Spent to 31/3/11 Remaining prject capital Working capital, exploration and overheads Total to be Funded Funded By Cash and liquid investments *Debt drawdown Total Funding * intended debt drawdown, banking facilities are currently being finalised

Jun-11 Aug-11 Jun-11 Jun-11 Jun-11 Aug-11 Oct-11 Dec-11 Dec-11 Dec-11 Q1 2012 US$M 576 217 359 65 424 210 214 424

Mining Contract Mining contract awarded to Leighton Asia Aug-10 First trucks arrived on site Aug-10 Haul road construction start Oct-10 Overburden removal complete

Jun-11 First ore mined to stockpile Aug-11 Gold & Silver Processing Plant Bulk earthworks contract awarded Oct-10 First concrete pour Mar-11 Steel, mechanical & piping, & mill installation contract awarded to McConnell Dowell Jun-11 Award of electrical and instrumentation installation contract Jul-11 Grinding SAG mill components delivered to site Jun-11 Ore primary crusher delivery to site Aug-11 CIL processing tank erection complete Oct-11 Wet commissioning start Dec-11 Tailings Storage Facility (TSF) TSF construction contract awarded Nov-10

TSF Phase I complete for receipt of tailings Dec-11 Engineering Engineering and procurement complete Dec-10 Exploration & Resources New exploration drilling results Sep-10 Updated resource statement Oct-10 Updated reserve statement May-11

3. Key Points relating to the Tor Uluala September 2011 Mineral Resource Estimate 3.1 Deposit Style The Tor Uluala Deposit is a High Sulphidation Epithermal (HSE) deposit similar to the other Martabe deposits at Purnama, Ramba Joring and Uluala Hulu which are interpreted to lie within a north-south trending structural corridor. The Tor Uluala resource is composed of multiple mineralised zone, many of which outcrop at surface, with a strike length of over 1km. Mineralisation is open at depth, to the north and to the south. Mineralisation is dominantly hosted in steeply west dipping to subvertical breccia units within an andesite host sequence. Epithermal gold-silver mineralisation followed early-stage acid sulphate wall rock alteration that produced vuggy and massive silica zones. The silica alteration zones grade outwards to an advanced argillic alteration envelope dominated by quartz-alunite which in turn grades out to argillic (kaolinite-illite) alteration. The mineralising system was focused around the breccia and structural zones which are presumed to be conduits for the hydrothermal fluids. Oxidation often overprints the alteration and mineralisation system. Oxidation is highly variable, with minor to strong oxidation extending to depths greater than 100m in places at Tor Uluala, although fresh sulphides can also be observed close to surface. 3.2 Drilling The September 2011 model for Tor Uluala used 81 diamond drill holes for 17,554m to define the resource. The majority of core was HQ sized with PQ in the upper 50-100m of most drill holes. No core was drilled smaller than HQ size. 3.3 Drilling Quality Assurance All of the diamond drill hole collar locations were surveyed using Total Station once completed. Downhole surveys were completed by Eastman camera with initial measurements at 20m from the collar with subsequent measurements at nominal 50m intervals and at the end of the hole. Drill hole locations and orientations were validated by plotting collar pick-ups against the drill design. Core recovery is generally excellent for all core sizes, averaging approximately 95% recovery. 3.4 Topography Surface topography was derived from an Airborne Laser Survey (LIDAR) conducted in June 2010 and sub-sampled to a 5x5m grid for use in 3D software. 3.5 Geological Logging All drill core was geologically logged for lithology, oxidation, alteration, structure, core recovery and RQD onto hard copy log sheets using a comprehensive set of logging codes.
7/9

3.6 Sampling Drill core was halved and sampled typically at 1m intervals although varying intervals between 1m and 3m exist within the database. Zones of silica alteration and quartz veining determined during geological logging were selectively assayed. All samples are of diamond core. 3.7 Analytical Methods All samples were analysed by Intertek Mineral Services (ITS), an internationally recognised company with offices and assay facilities in Jakarta and a sample preparation facility in Padang, South Sumatra. Samples were analysed for Au, Ag, Cu, As and Sulphide Sulphur. Samples were trucked from site to Padang. At the Padang facility samples were crushed to 10mm, then split to approximate 1.5kg samples. These were then pulverised to -200mesh and split to a 250 gram sample then sent to ITS Jakarta for assay. Assaying techniques employed are given in Table 3 below. Table 3: Assay techniques employed for the Tor Uluala samples. (SCIS is used to analyse the percentage of sulphide sulphur)

3.8 Assay Quality Control and Quality Assurance PT Agincourt Resources (PTAR) has a suite of Certified and non-certified Standards (Standards) covering a range of grades and elements (including Au, Ag and Cu but not Sulphide Sulphur). Two hundred and six (206) Certified Standards, sourced from Geostat Pty Ltd and Ore Research and Exploration (OREAS) Pty Ltd, were submitted as part of this campaign. 8/9

Either a Standard or Blank was inserted at the rate of one in every 20 samples. Overall ITS performed very well with these standards, with the few anomalies observed considered likely due to mislabeling or data mismatching errors. 3.9 Bulk Density Bulk density was determined on all holes at regular intervals using the Archimedean method utilising plastic wrap (gladwrap) sealing of oven dried drill core samples (0.2m lengths) samples. Average bulk densities were assigned in the block models according to the lithologies, oxidation and mineralised zones. Core specimens for determination were selected by the geologists based on encompassing the range of lithology and alteration types. In general a determination was taken every 10m, if lithology and alteration is relatively homogeneous in that interval or more frequently if required. A total of 1,214 bulk density determinations were available for the estimate. Three Bulk Density Standards have been routinely used to verify the bulk density measurements. These Standards are based on PQ, HQ and NQ core sizes manufactured from concrete. A measurement of one of them was taken approximately for every 10 routine measurements prior to drill hole APSD595, and for approximately every five thereafter. 3.10 Interpretation Geological modeling was carried out using Micromine software and on paper cross sections. The definition of the Tor Uluala mineralisation interpretations by PTAR were guided by the underlying lithology, alteration and structure interpretation as well as the distribution of grades. The importance of lithology, alteration and structure in controlling the mineralisation is well recognised and these features were utilised in guiding the Tor Uluala mineralisation interpretations. The mineralisation boundaries are mostly gradational, depending upon plumbing network & amount of silica alteration. Generally the gold content is proportional to the degree of silica alteration. PTAR supplied Cube Consulting with a set of sectional interpretations based upon a 0.5 g/t delineating cut-off for gold and 2.5 g/t delineating cut-off for silver, along with lithology, alteration and structural considerations. The interpretations were created using 5m downhole composites to assist with interpreting continuity. Cube subsequently performed some minor modifications to maintain consistency and continuity between sections and to ensure wireframes honoured the 2m composites used by CUBE. The final interpretation and domaining was reviewed and accepted by PTAR site geologists. This interpretation resulted in 12 distinct wireframed gold domains numbered 10 to 21, and ten wireframed silver domains numbered 30 to 39. Not all of the mineralisation was able to be wireframed due to discontinuity between sections without obvious down dip or along section connection. The un-wireframed intercepts were incorporated into two additional alteration domains that surround the mineralisation domains; 100 (inside the advanced argillic and silicic alteration domains), and 101 (inside the argillic alteration domain). 9/9

3.11 Grade Estimation Cube utilised Ordinary Kriging to estimate gold, silver, copper, sulphide sulphur, arsenic and into a 3D block model. Relative variogram models, and search neighbourhoods were used to interpolate high cut composite data. Quantitative kriging neighbourhood analysis was initially used to establish optimal estimation parameters for the gold domains in some less informed regions and to moderate some smoothing effects of the ordinary kriging process. All block estimates were based on interpolation into 25m N x 12.5m E x 10m RL parent cells, sub-celling to 6.25m N x 3.125m E x 2.5m RL where necessary. Top cuts to gold and silver were applied based on univariate statistics and spatial plots for each domain. Gold top cuts were applied to 20 samples and range from 0.5 ppm to 10 ppm depending on domain. Silver top cuts were applied to 70 samples and range from 7 ppm to 70 ppm depending on domain. 3.12 Cut-off Grade A lower gold cut-off grade of 0.5 g/t Au was used for reporting. 3.13 Mineral Resource Classification Mineral Resources were classified based on data density, confidence in interpretation, number of samples involved in the estimate, confidence in the input data, and structural and grade continuity of the mineralised zones. Based on these criteria, all material in this resource estimate has been classified as Inferred. Reporting is not constrained in depth or lateral extent by a nominal optimisation shell or other financial constraint. However the resource presented is considered to have reasonable prospects of eventual economic extraction. Additional drilling and mining studies are required to gain further confidence in the continuity of the mineralisation, particularly at higher cut off grades. 3.14 Block Model Validation Mineral resource estimates were validated against the original data. This was done by visual inspection of the estimated block models against the original data, and by statistical analysis where the global input and output statistics and semilocal input and output statistics were compared with each other.