Phase 3 drilling beneath the 2011 intercepts has been successful in extending the gold mineralization to a depth of about 250 meters without significant loss of gold grade or mineralized widths. In-fill drilling at or above this level to establish continuity and to bring drill hole spacing to acceptable levels has also been largely successful with only occasional low grade (DDH-101) intercepts within the higher grade envelope.

Deeper drilling, in the western portion of the 55 Zone has encountered changes in lithologies in the areas where the veining was projected which appear to inhibit vein formation and gold deposition at the depths between 250 to 525 meters below surface. The brittle intrusive rocks which appear to host the veining and gold preferentially are replaced at these depths with basalt and structures here are generally limited to deformation and sheared zones with only minor gold values (DDH-113,-125,-130,-136,and-148). This situation improves markedly towards the central and eastern portions of the deposit with positive results from holes DDH-106, DDH-118, DDH-141 and DDH-065. Visible gold was noted in holes DDH-154 and DDH-165 at intercepts 300 meters and 375 meters below surface respectively. It also appears that the favourable intrusive host rocks persist at depth to the east beyond their surface expression. The 55 Zone demonstrates a trend to the east with depth both in gold values and favourable lithology and structure which, now recognized, will guide further drilling. This trend is shown in the polygons in the longitudinal section below. The veining in the very deep drill hole DDH-127 and the wedged intercepts from it drilled in the western portion are also located in intrusive rocks indicating that the less favourable basalts encountered in the west may not persist to depth and that the quartz veining and favorable lithology, while not auriferous in this area, persist to very great depths.

To view map please click on the following link: http:media3.marketwire.comdocsROGmap.pdf

EXPLORATION ACTIVITIES IN OTHER AREAS OF YARAMOKO

Core drilling has been carried out in target zones identified through previously announced reverse circulation drilling away from the 55 Zone. Assays are pending for 4 holes, comprising 891 meters, at the 109 Zone and 12 holes, comprising 3,584 meters, at the 117 Zone. Drill intercepts and assays will be reported once assays for these suites of holes are received.

Bagassi South has been the subject of two exploration campaigns in the past, Roxgold Inc. in 2010-2011 and a previous operator in 2007. Drill results from these programs yielded numerous significant gold intercepts but follow up had been deferred due to the high grade results obtained from the 55 Zone. Work on the Bagassi South property, approximately 2.5 kilometers south of the 55 Zone, has been ongoing through 2012. Work during this year has included 375 soil samples, 5,354 meters of RAB drilling in 379 holes, 14 reverse circulation holes comprising 2,224 meters and 4 core holes comprising 2,380 meters. Of the more than 4,000 samples sent for analysis to in-country accredited assay labs, assay results have been received for only the soil survey and some of the RAB holes. It is felt that, given the gold resource established now at the 55 Zone, a strong effort should be made to try and discover new gold zones in close proximity to potentially provide additional tonnage.

MINERAL RESOURCE ESTIMATE PARAMETERS AND METHOD (as provided by AGP)

—  The block model mineral resources were estimated for the 55 Zone of the

    Yaramoko Property. The estimate encompasses the main vein and a smaller

    accessory vein located on the footwall of the main vein. The estimate

    was completed based on the concept of an underground operation. No other

    zones on the Yaramoko property were evaluated.

— The three-dimensional (3D) wireframe model was based on lithology
typically taking into account grade in excess of 2.0 gt with exception
made for continuity and to account for a 1.5 meter horizontal minimum
mining width. The 3D wireframe describes the shape of the mineralized
horizon. The grade boundaries with the hanging wall and footwall waste

are both sharp.

— Preliminary metallurgical work is underway and no results were available
in order to assess potential mill recoveries.

— A (3D) geological and block model was generated using GEMS(C) software.
The block model matrix size of 2x5x5 metres (width x length x height)
was selected with consultation with the engineering team from AGP and
was based on the size that was deemed suitable for an underground narrow
vein mining scenario.

— Out of 135 drill holes intersecting the mineralization, 7 holes have
assays pending and were removed from the estimate leaving the model to

be interpolated with the remaining 128 holes totaling 31,775 meters of

drilling completed by Roxgold from mid-2011 through 2012.

— All drill holes are diamond drill core with the majority of the samples
collected and assayed within the mineralized zone between 1.0 to 1.4
metre sample intervals.

— The composite interval selected was 1.5 metres down hole with composite
remnants backstitched to the previous interval.

— Densities were determined from 456 representative rock samples using
industry standard methods. For the Main zone a density of 2.74
tonnesmetre3 and 2.72 tonnesmetre3 for the small Footwall zone was
applied to the model using the density of each lithology within the 55

and 55 Footwall Zone.

— Variography studies within the Main zone indicated a long axis pointing
in the strike direction of the deposit and a short down-dip axis. The
result of the variography was likely more indicative of the drilling
pattern than the true trend of the deposit.

— For the treatment of outliers, raw assays were capped to 280 Au gt in

combination with a search restrictions applied on composites values
greater than 50 Au gt. The procedure used allows the deposit to retain
the high grade assays while limiting their influence during the
interpolation to a maximum of 15m x 25m x 25m (width x length x height).

— Inverse distance cube was used for the grade interpolation with a
nearest neighbour check model.

— The interpolation was carried out in multiple passes with increasing
search ellipsoid dimensions. Classification for all models was based
primarily on the pass number followed by an adjustment to the class
model, based on a diamond drilling density (core area) and the distance
to the closest sample.

— No mining plans have yet been prepared for the deposit however from the
geometry of the deposit, it seems likely that shrinkage, cut and fill or
long hole stoping mining methods, with or without backfill, may be
considered for future extraction.

— Mineral resources that are not mineral reserves do not have demonstrated
economic viability.

— Under CIM definitions, Mineral Resources should have a reasonable
prospect of economic extraction. In order to assess the Mineral
Resources an insitu resource cut-off grade of 2.0 gt gold is
recommended. A gold price of US$1,100 per troy ounce was used for the
estimate.

— The quantity and grade of reported inferred resources in this estimation
are conceptual in nature and there has been insufficient exploration to
define these inferred resources as an indicated or measured resource and
it is uncertain if further exploration will result in upgrading them to
an indicated or measured resource category.

— Rounding of tonnes as required by reporting guidelines may result in
apparent differences between tonnes, grade and contained metal content.

— A number of collar positions were validated during the site visit using
a hand held GSP. Assays were validated against the original certificates
obtained directly from the issuing laboratories. Independent character

samples collected on-site all confirm the presence of gold on the
property.