Exploration review

In 2006, exploration activities concentrated on the extension of known orebodies and the discovery of new orebodies both at producing mines and exploration sites. In addition, the company continued with its expansion of the African Footprint strategy within the most prospective gold belts of both West and East Africa and has operations in six African countries boasting a portfolio of 128 targets on 18 526km2 of groundholding.

At Loulo, the company continues to build its resource base, which is now over 10 million ounces. As well as the resource conversion work, exploration is focused on the next discovery by systematically evaluating a portfolio of targets within the lease area. Of these, the Faraba and Baboto districts are showing the most promise.
At Morila, the next discovery continues to evade detection as plus 40 000 metres of regional drilling have been completed. However, the low grade footprint continues to be defined. Research points towards an intrusive related ore genesis to the deposit.
In south Mali, a regional gravity survey is enabling the identification of targets for drilling in 2007.
In Senegal, a 10 000 metre Rotary Air Blast ("RAB") drilling programme has commenced to evaluate 12 targets. This will help in prioritising additional targets to Delya, Bambaraya and Sofia for diamond drilling later in the year.
In Burkina Faso, drilling is defining a broad, low grade mineralised system over a three kilometre strike at Kiaka. Elsewhere in the company's permit portfolio, regional programmes are identifying targets for follow-up work in the coming year.
In Ghana, exploration activities recommenced on a portfolio of four permits. First pass regional exploration programmes have been completed and returned positive results from one of these, Bole, which will see follow-up programmes in 2007.
In Tanzania, generative work is the driver to build a new portfolio of projects after Kiabakari did not meet the company's criteria for further investment.
In the Côte d'Ivoire, exploration activities successfully recommenced with a tactical diamond drilling programme at the Tongon prospect, situated within the Nielle permit, in the north of the country. Preparations are now almost complete to start a ~30 000 metre feasibility diamond drilling programme in 2007.
A generative team has been established to re-assess the African countries within which Randgold Resources is not currently operational, to identify potential value adding projects that would be of interest to the company.

In summary, the company has a quality portfolio of exploration projects in both West and East Africa. This reflects the company's business strategy of organic growth through exploration and its primary objective to build sustainable mining projects with significant returns. This strategy is attested to by its discovery and development track record, which includes the Morila and Loulo mines, both in Mali, and the plus three million ounce Tongon project, currently at the feasibility stage in the Côte d'Ivoire.

MALI

LOULO
In addition to the resource conversion work and the deep drilling at both the Yalea and Gara orebodies, exploration has concentrated on the next discovery by evaluating a quality portfolio of targets within the 372km2 permit area.

At the Gara deposit, new information from the continual excavation of the deposit and re-logging and re-interpretation of diamond drill core established that high grade mineralisation is associated with the 0300/0400 trending axial planes of quartz tourmaline folds and is concentrated in the hinge zone of a large overturned antiform.

Six diamond drill holes for a total of 3 809 metres have confirmed the geological model and intersected strong quartz tourmaline alteration and mineralisation up to 500 metres south of the existing wireframe; L0CP117: 5.80 metres at 3.18g/t from 333 metres and L0CP118: 4.76 metres at 0.64g/t from 409 metres and 6.15 metres at 1.39g/t (including 1.20 metres at 3.70g/t) from 459 metres; LOCP120: 5.80 metres at 9.16g/t (including 2.90 metres at 16.66g/t) from 757.4 metres; LOCP124: 7 metres at 17.95g/t from 551.75 metres (including 1.80 metres at 59.36g/t). LOCP126: 9.20 metres at 1.02g/t from 578.60 metres and 6.40 metres at 10.37g/t from 591.80 metres; LOCP125: 6.15 metres at 11.22g/t from 820.35 metres. Follow-up diamond drilling will further evaluate the potential of this target down plunge.

Loulo: Grade Model of Gara Orebody with Diamond Drill Holes to the South of the Existing Wireframe

The Baboto structure is part of a plus 5 kilometre mineralised structure which hosts the known targets of Baboto South, Central and North. At Baboto South a 23 hole, 2 368 metre RC drilling programme was completed over a strike length of 1.3 kilometres. The results returned a continuous zone of mineralisation over the entire strike length with an average width of 15.63 metres and grade of 1.82g/t to vertical depths of 95 metres.

Gold mineralisation is associated with both massive and disseminated pyrite and is hosted mainly in silica-carbonate altered sandstones between hangingwall and footwall shears.

At Faraba, 48 RC holes for 5 573 metres and eight diamond holes for 2 248 metres have been completed during the year testing a three kilometre strike length of an overall ten kilometre anomalous corridor. The drilling has currently identified two pods of gold mineralisation; the north zone and the main zone. These two zones are connected by zones of narrow alteration and mineralisation which have been dissected and offset by complex faulting and are not fully understood at this stage. The north zone has a strike length of 600 metres, an average width of 12 metres, an average grade of 2.20g/t and has been tested to vertical depths of 230 metres.

The main zone is a 100 metre wide anomalous structure which currently measures 400 metres in strike length and has been tested to a vertical depth of 180 metres. At its core is a 350 metre long mineralised zone with an average intersection width of 43 metres and an average grade of 2.60g/t resulting in an inferred resource of 567 000 ounces.

Loulo Permit: Portfolio of Gold Targets

This zone contains a high grade unit whose width and grade averages 5.50 metres at 8.70g/t over the 350 metre strike length. Drilling within this zone has confirmed the presence of extensive silica-carbonate alteration within coarse sediments and grits.

Trenching and drilling programmes will be completed to test the 10 kilometre regional Faraba target, where 14 additional dilation zones have been mapped, as well as to further define the north and main zones of mineralisation during 2007.

At P64, a 1.5 kilometre plus 100ppb north-northwest soil anomaly characterises the target. To date only 300 metres of strike has been drill tested with seven RC holes for 1 097 metres and five diamond core holes for 1 396 metres. The results highlight a 200 metre strike of mineralisation which includes: P64RC03: 7 metres at 1.47g/t and 4 metres at 1.03g/t, P64RC04: 29 metres at 1.03g/t, P64RC07: 5 metres at 2.31g/t, P64DH01: 28 metres at 1.72g/t, 16.2 metres at 4.41g/t and 11.80 metres at 1.44g/t and P64DH02: 5.80 metres at 3.43g/t and 8.20 metres at 6.51g/t. Mineralisation is associated with chlorite and magnetite rich sediments plus localised bands of quartz tourmaline and is bounded by a hanging wall and footwall shear which trend north-south and dip sub-vertically.

To increase the company's knowledge and understanding of the Loulo mineralised district, a three year PhD research thesis is being undertaken in conjunction with Kingston University, London, England.

SELOU
The Selou area consists of three permits directly south of the Loulo exploitation lease and work to date has identified two main targets; Sinsinko and Boulandissou.

The Sinsinko target is underlain by a 1.75 kilometre northsouth trending plus 30ppb gold soil anomaly; follow-up trench and RAB drilling have confirmed a bedrock source to mineralisation. One diamond hole SND001 drilled below trench BET05: 76 metres at 0.83g/t returned 5.70 metres at 1.71g/t, 4 metres at 3.11g/t, 1.80 metres at 2.77g/t and 4.20 metres at 7.10g/t, all within a 21 metre mineralised envelope. The mineralisation is associated with tourmaline-sericite-carbonatehaematite- silica alteration of a greywacke host. Further drilling is planned in 2007.

Boulandissou is an 8 kilometre anomalous shear corridor. A reconnaissance diamond hole BND01 drilled beneath trench BNT02: 28 metres at 3.31g/t returned 6.60 metres at 1.28g/t.

Gold anomalism and mineralisation in the diamond hole is hosted in altered quartzite and associated with weak disseminated pyrite and arsenopyrite. The alteration patterns are dominated by brecciation in association with silica, tourmaline and weak sericite. Additional trenching and drilling will be completed during 2007 to evaluate this target.

MORILA EXPLOITATION LEASE
An exploration strategy was developed with the primary aim of providing an assessment of the full resource potential within the greater Morila lease area. The flat geometry of a Morila style orebody, especially as a "blind or buried" exploration target, prevents the use of conventional exploration geochemical or geophysical target generation techniques. In the search for a buried 'Morila' the only known viable, primary targeting tool available is a regularly spaced deep diamond drilling programme, to date a total of 89 holes for 44 045 metres have been drilled, to an average depth of 500 metres. The results, although failing to identify an additional deposit have so far intersected the low grade (plus 0.10g/t) footprint to the deposit. This data will be used as a vector for follow-up holes to target for higher grade mineralisation.

A six month post doctoral research project has been set up through the Geology Department of the Australian National University and the Research School of Earth Sciences. The aim of this study is to understand the relationship of gold mineralisation to high grade metamorphism and partial melting of the surrounding sediments. Preliminary results indicate an ore genesis related to the contact thermal aureole of an igneous intrusive.

MORILA REGION AND SOUTHERN MALI
In the Morila region, Randgold Resources has been exploring on its permits, immediately adjacent to Morila mine, for several years. The techniques used to date have been successful in locating mineralised structures such as Ntiola and Kona. However, so far no economic mineralisation resembling the Morila deposit has been located outside the mine-lease. Despite the lack of success in the area, Randgold Resources will continue exploring until all reasonable avenues of research have been investigated. An indication of this commitment is a 3 000 metre diamond drilling programme which was completed before the annual wet season in June.

As few gold targets at surface remain untested and rock outcrop only accounts for 0.5 percent of the surface regolith, this programme was designed to provide geologists with geological and structural information to assist with interpretation in this most difficult of terrains. This drilling programme has allowed the team to modify its geological model for the area (which was based on sparse information) with the aim of identifying broad, favourable areas for the hosting of blind, Morila-type mineralisation. To complement the data from this drilling programme a 40 by 40 kilometre ground gravity survey has been completed, centred over the Morila deposit. The results return an arcuate anomaly immediately south of the deposit and both to the northeast and northwest of the grid. Modelling is attempting to estimate the depth and type of geological body responsible for these results.

Randgold Resources' joint ventures with Japanese company OMRD have been terminated following a full evaluation of the Kekoro, Sagala, Diamou and Seriba-Sobara permits, which did not meet the company's criteria for further investment. In Southern Mali, Randgold Resources' regional teams continue to identify opportunities and apply for new ground. This is in line with the long term strategy of turning over ground to supply the base of the resource triangle with quality targets.

Mali West and South: Randgold Resources Exploration Licenses

SENEGAL

The Senegal portfolio consists of five permits covering 1 421km2, located within the Sabodala volcano-sedimentary belt in the east of the country. The company has made progress throughout the field season in evaluating its portfolio where the emphasis has been on developing targets to the drill stage. A 10 000 metre RAB drilling programme has commenced to test 12 targets. This will help in prioritising additional targets to Delya, Bambaraya and Sofia for diamond drilling later in the year. The year's field programmes also resulted in the rejection of 14 targets.

Delya was a new target identified during 2006 and locates on the Main Transcurrent Shear. This is a major terrain boundary structure between the greenstone belt and adjacent sedimentary basin. The target was identified by soil geochemistry, which defined a 6 kilometre by 100 metre plus 20ppb gold anomaly. Ten follow-up trenches have identified two parallel zones of gold mineralisation over a 1 kilometre strike, the best results of which are; DLT003: 11.15 metres at 9.60g/t; DLT004: 4 metres at 1.60g/t; DLT005: 4.5 metres at 7.54g/t; DLT006: 7.45 metres at 1.98g/t and 9.90 metres at 4.96g/t; DLT008: 18 metres at 0.60g/t; DLT009: 2 metres at 5.39g/t and DLT010: 2.5 metres at 0.80g/t. The trench results have been tested at depth by five diamond core drillholes (1 000 metres); DLDD001: 9.81 metres at 1.80g/t; DLDD002: 12.40 metres at 5.10g/t; DLDD003: 3 metres at 1.80g/t; DLDD004: 3.80 metres at 4.80g/t and DLDD005 returned no mineralised intersection. Gold mineralisation is hosted within a package of schists, strongly sheared, affected by silica-sericite-irongraphite alterations. Sulphides are present as disseminated pyrite and arsenopyrite.

Senegal: Prioritising Drill Targets

At Sofia, work was concentrated in collating all the previous exploration data, as no drilling was completed in 2006. Sofia is part of a seven kilometre anomalous north-south structural corridor which also hosts the Mikona, Mativa and Matiba targets within ground held by Randgold Resources. This system continues to the north for an additional 10 kilometres and hosts the Niakafiri deposit and Sabodala deposit. RAB drilling is planned in the gap areas to try and link all the Sofia, Mikona, Mativa and Matiba targets. So far 3.4 kilometres of strike have been tested by drilling where results return both broad low grade (44 metres at 2.00g/t) mineralisation and narrow high grade (6 metres at 9.50g/t) intercepts. At present, the interhole spacing is 400 to 600 metres which will be infilled during the next round of drilling. In the meantime RAB drilling will test the Mikona, Mativa and Matiba targets within the Sofia corridor and the gaps in between.

At Bambaraya, trenching and early stage drilling has defined two sub-parallel zones of mineralisation at surface, over a strike length of 1 kilometre. The best trench intersections returned are; BBTR001: 13.20 metres at 3.59g/t, BBTR002: 18 metres at 2.93g/t, BBTR003: 8 metres at 4.50g/t, BBTR004: 12 metres at 4.06g/t and 4 metres at 5.48g/t, BBTR006: 14 metres at 2.01g/t and 9.5 metres at 1.13g/t. BBTR010: 16 metres at 1.70g/t and BBTR007: 18 metres at 2.26g/t. To date only three diamond drill holes have tested this zone; with BBDDH002 returning the best intersection: 12 metres at 3.17g/t.

Mineralisation is hosted within northeast trending pillow basalts and is associated with silica-sericite-tourmaline-iron carbonatepyrite alteration. Exploration work has been completed to the northeast of the known mineralisation where a felsic intrusive has intruded the structure.

BURKINA FASO

In Burkina Faso the company made excellent progress over the last year:

The development of a regional geological model.
The consolidation of a plus 2 000km2 permit portfolio in the south of the country, covering the southern extension of the Markoye fault and associated splays.
The completion of the first phase of regional exploration across all the permits, which has identified a strong base of targets to the resource triangle.
The identification of an advanced target, Kiaka, which has broad zones of low grade mineralisation over a 3 kilometre strike length.
The development of an all Burkinabe exploration team.

Burkina Faso - Drilling the Kiaka Permit

On the Kiaka permit a 9 040 metre, 549 hole RAB programme was completed testing three targets within the Kiaka permit. The most advanced of these targets is Kiaka which also saw a 1 125 metre, 11 hole RC drilling and a six hole, 1 371 metre diamond core programme completed testing a three kilometre long mineralised system.

Two styles of mineralisation were identified from the drilling at Kiaka:

A narrow high grade hangingwall zone; the mineralisation is associated with strongly altered, (silica-biotite-chlorite) sulphide rich (5-22% arsenopyrite) sediment.
A broad low grade main zone; the mineralisation is associated with schists and quartzites and fine disseminated sulphides (pyrite).

These mineralised zones locate within a northeast trending, belt parallel, regional shear. Sinistral reactivation of north-south belt discordant structures has created dilational openings along the northeast regional shear and therefore preferential sites for mineralisation.

The results of the drilling at Kiaka have identified a 3 kilometre long mineralised system, within which locates a 1.20 kilometre continuous zone of gold mineralisation, 800 metres of which is 100 to 200 metres wide grading 0.80 to 1.60g/t and has been drill tested to vertical depths of 200 metres; it is open in all directions. The best intercept returned to date is from KDH05: 73 metres at 2.14g/t (from 41 metres) including 11 metres at 3.40g/t (from 44 metres) and 14 metres at 3.00g/t (from 58 metres). Infill and step out drilling are planned to commence in February 2007.

In addition, first pass regional soil geochemistry has been completed over the portfolio of nine permits. The results delineated 11 regional, five identified and two follow-up targets.

GHANA

In Ghana, the company has made good progress with the completion of first pass regional exploration programmes over our portfolio of four permits; three of these permits are in the process of being relinquished following unprospective results while one is progressing with positive results. New applications are being made in discussions with the Minerals Commission.

On the Bole Northeast permit, an 800 metre by 100 metre regional soil sampling programme has returned positive gold results.

A 25 kilometre long, northeast trending (060) anomalous corridor has been defined at the contact between granite and volcaniclastic sediments within which locate two target areas.

Zamsa consists of two sub parallel anomalies; the first is a plus 20ppb gold in soil anomaly measuring 14 kilometres long by up to 2.5 kilometres wide. This anomaly is associated with the main regional Bole-Bolgatanga shear and locates at the contact between granite and volcanoclastic sediments. The second is a plus 10ppb gold in soil anomaly measuring five kilometres long by one kilometre wide and is coincident with the axis of a regional airborne magnetic and electro magnetic anomaly within the volcaniclastic sediments.
Navrongo locates five kilometres to the northeast of Zamsa along the same Bole-Bolgatanga shear. The 'gap' relates to a national forest which cannot be permitted.
Ground truthing of anomalies indicates a gentle undulating topography with residual soils and poor outcrop. Test pitting is in progress to understand the regolith profile and mobile element dispersion prior to infill soil sampling.

Ghana: First Pass Regional Exploration Programmes Returned Positive Results

CÔTE D'IVOIRE

Apart from continuing to add resource ounces at Loulo, the big success story of 2006 was resuming exploration activities in the Côte d'Ivoire, with the completion of an eight hole, 1 992 metre tactical diamond coring programme at the Tongon prospect, which locates in the Nielle permit, in the north of the country.

Geologically the Tongon area comprises a sub-vertical northeast to east-northeast trending package of intercalated clastic and mafic to intermediate volcano-sedimentary lithologies. These lithologies have been intruded in the central and northern regions by large oval shaped northeast trending granodiorite to quartz diorite igneous bodies. In the south, diorite dykes and small micro-gabbro bodies have intruded along east northeast to east trending structures. Mineralisation locates in two zones; northern and southern.

Five holes were completed on the northern zone, further testing a 1.5 kilometre strike length of the main shear.

Three holes were completed in the southern zone to confirm the results of the prefeasibility study; the results are presented below.

The overall geometry of the mineralised zones suggests a significant component of dextral shear which is compatible with previous interpretations. Dilation occurs along east northeast orientated fault segments which appear to have been reactivated during mineralisation. The main structural direction is northeast, representing belt parallel shearing, together with the presence of north-south belt discordant structures. The interaction between these two structural trends, in this case primary dextral movement along north-south trending faults, causes reaction of the northeast belt parallel structures and in turn creates dilation on the east-northeast structures.

The 30 000 metre diamond coring feasibility drilling has commenced.

Cote d'Ivoire: Tongon Prospect

TANZANIA

In Tanzania, generative work is the driver to build a new portfolio of projects; we are relooking at the Southern Lake Victoria Goldfield, the Proterozoic mobile belts and new greenstone belts within the Craton.

An evaluation of the old Kiabakari mine was completed; RC and diamond core drilling tested the immediate vicinity of the old mine and a 2 000 metre strike along the Kiabakari fault. Anomalous values were returned along a 1 800 metre segment of the fault associated with three structures within a 300 metre wide corridor. However the main mineralisation is constrained to a 500 metre long by 35 metre wide zone grading 2 to 6g/t, within which the high grade payshoot, measuring approximately 100 metre by 30 metre grading plus 6g/t has been mined out to 350 metres below the surface and selectively to 500 metres below the surface.

Randgold Resources has fully tested the Kiabakari target and concludes that there is no big mineralised system which fits its criteria for further investment. This project has been handed back to the government.

In the Mara Region, mapping and trenching have tested the Mara fault corridor which has been traced for 3 kilometres within the Nyabigena South permit. This fault system hosts the Nyabirama mine (1 million ounce resource), the Komarera deposit (400 000 ounces) in addition to two old colonial mines (Mara and Pontama).

The results are encouraging with two 15 to 30 metre anomalous zones, within a 70 to 90 metre wide east-west structural corridor, locally mylonitic. Geologically the target is underlain by granodiorite gneisses with highly deformed and carbonated mafic dykes. The main alterations observed are sericite-silicapyrite. Trench results from west to east include:

NYTR17: 34 metres at 0.16g/t and 24 metres at 0.51g/t, including 14 metres at 0.82g/t.
NYTR21: 16 metres at 0.10g/t and 18 metres at 1.10g/t.
NYTR31: 22 metres at 0.61g/t and 20 metres at 0.10g/t.

An RC drill programme will follow-up these results in 2007.

More regionally, three Barrick permits in the Singida-Dodoma region were incorporated into the current joint venture agreement with Randgold Resources.