Roxgold Increases Indicated Mineral Resources by 97% to Over 1 Million Ounces at Séguéla; Reports Maiden Inferred Mineral Resource at Koula of 281,000 oz at 8.1 GPT AU

Roxgold Inc. is pleased to report an updated Mineral Resource Estimate for the Séguéla Gold Project (“Séguéla”) located in Côte d’Ivoire. The updated Séguéla NI 43-101 Mineral Resource estimate includes an additional 56,600 m of Reverse Circulation (“RC”) and diamond core (“DD”) drilling since the completion of the Séguéla PEA in April 2020. The drill program prioritized infill drilling of Antenna, Ancien, Boulder and Agouti to increase resource confidence, while rapidly advancing the high grade Koula discovery to its maiden Inferred Resource.

Total Indicated Mineral Resources increased 97% to 1,044,000 ounces (“oz”) of gold grading 2.5 grams per tonne (“g/t”) since the Preliminary Economic Assessment (“PEA”) was released in April 2020 (refer Company press release dated April 14, 2020). Total Inferred Mineral Resources are estimated at 370,000 oz at 4.8 g/t, due to the contribution of a maiden Mineral Resource estimate from the recently discovered high grade Koula prospect, which returned Inferred Mineral Resources of 281,000oz at 8.1 g/t.

Highlights:

• Total Indicated Mineral Resources: Increased by 97% to 1,044,000 oz at 2.5 g/t from 529,000 oz at 2.3 g/t previously.
• Koula deposit: Maiden Inferred Mineral Resource of 281,000 oz at 8.1g/t.
• Antenna deposit: Increase in Indicated Mineral Resource to 586,000 oz at 2.2 g/t and an Inferred Mineral Resource of 69,000 oz at 1.9g/t. This compares with the previous estimate carrying an Indicated Mineral Resource of 529,000 oz at 2.3 g/t and Inferred Mineral Resource of 64,000 oz at 2.2 g/t.
• Ancien deposit: Upgraded Mineral Resource with an initial Indicated Mineral Resource of 250,000 oz at 5.4g/t and an Inferred Mineral Resource of 11,000 oz at 10.6g/t. This compares with the previous estimate carrying an Inferred Mineral Resource of 261,000 oz at 6.1 g/t.
• Agouti deposit: Upgraded Mineral Resource with an initial Indicated Mineral Resource of 111,000 oz at 2.4g/t and an Inferred Mineral Resource of 6,000 oz at 1.8g/t. This compares with the previous estimate carrying an Inferred Mineral Resource of 110,000 oz at 2.6 g/t.
• Boulder deposit: Upgraded Mineral Resource with an initial Indicated Mineral Resource of 97,000 oz at 1.7g/t and an Inferred Mineral Resource of 3,000 oz at 1.2g/t. This compares with the previous estimate carrying an Inferred Mineral Resource of 72,000 oz at 1.2 g/t.

“It has been an exceptional year at Séguéla as we have continued to demonstrate the potential for the project to grow in quality and scale,” stated John Dorward, President and CEO. “The significant increase in Indicated Mineral Resources to over one million ounces provides the foundation for a robust Feasibility Study which will also include the very high grade Koula deposit with a maiden inferred resource of 281,000 ounces at a remarkable 8.1 g/t giving it the potential to be one of the highest grade open pits in west Africa. We currently have 4 rigs at Koula undertaking infill drilling and exploration drilling as the deposit remains open along strike and at depth.

“Since acquiring Séguéla in April 2019, we have seen the project grow at a rapid pace from the 430,000 ounces of inferred Mineral Resources at acquisition to over a million ounces of Indicated and an additional 370,000 ounces of Inferred Mineral Resources today. We continue to believe we have only begun to tap the ultimate potential of the Séguéla Project and are eager to continue to uncover and test the wealth of additional targets present on the property. While our accomplished exploration team continues their work at Séguéla, the critical path for the Séguéla project plan is on track with the Feasibility Study scheduled for the first half of next year, followed soon thereafter by a construction decision towards the goal of achieving first gold pour at Séguéla in 2022.”

Table 1 – Séguéla Mineral Resource Statement Summary

The resource estimate announced today is an update from the resource estimate used in the PEA disclosed by the Company on April 14, 2020. In the proposed mine plan outlined in the PEA, a total of 890,000 oz from all resource categories are mined from the Antenna, Ancien, Boulder and Agouti deposits. The updated resource estimate does not affect the validity or currency of the PEA. With the increase in Indicated Mineral Resources from Antenna, Ancien, Boulder and Agouti and the maiden Inferred Mineral Resource at Koula, Roxgold believes there is the potential for increased mine life at Séguéla and higher potential future production, which will be assessed in future studies.

Indicated Mineral Resources increased 97% to 1,044,000 oz Au from 529,000 oz Au relative to the April 14, 2020 estimate (refer Company press release dated April 14, 2020). Inferred Mineral Resource decreased by 27% to 370,000 oz Au from 508,000 oz Au relative to the April 14, 2020 estimate reflecting the conversion of prior Inferred Mineral Resources to Indicated Mineral Resources and the addition of the new high-grade Koula deposit.

The Mineral Resources and Mineral Reserves models were prepared in conformity with The Canadian Institute of Mining, Metallurgy and Petroleum’s (CIM) Estimation of Mineral Resources and Mineral Reserves Best Practices Guidelines (November 2019) and are classified per the CIM Definition Standards for Mineral Resources and Mineral Reserves (May 2014).

Table 2 – Séguéla Mineral Resource Comparison^1,2

Koula

Located approximately 1km to the east of Antenna, Koula was discovered through field reconnaissance and coincident recent artisanal workings in an area previously considered to be a lower exploration priority. An initial 10 hole reconnaissance RC drill program successfully returned intersections of high grade mineralization extending over a projected strike of more than 300 metres across 5 drill lines, with the deepest intersection approximately 150 metres below surface. This program was extended with 10 lines of drilling infilling and extending coverage to over 400m of strike and where the high grade mineralization remains open at depth.

Mineralization is hosted by quartz-carbonate veining associated with a well developed mylonitic fabric within and along the interpreted margins of a tholeiitic basalt which in turn has been tightly folded. Coarse gold is commonly recorded in the higher grade zones, with drilling highlighting a consistent moderate southerly plunge to the high grade core over at least 450m, and where it remains open to the south.

With four drill rigs operating at Koula, an infill program of RC and diamond drilling to 50m centres is ongoing to support the rapid advancement of Koula to Indicated status and inclusion in the upcoming Feasibility Study. Recent high-grade results from the conclusion of the Inferred drilling program include:

• 22 metres (“m”) at 16.4 grams per tonne gold (“g/t Au”) in drill hole SGRC928 from 131m including
  • 2m at 52.7 g/t Au from 135m and
  • 4m at 53.0 g/t Au from 138m
• 14 m at 14.6 g/t Au in drill hole SGRC968 from 40m including
  • 1m at 24.9 g/t Au from 40m and
  • 5m at 31.4 g/t Au from 44m
• 18 m at 9.5 g/t Au in drill hole SGRC965 from 21m including
  • 4m at 32.7 g/t Au from 26m
• 19 m at 3.4 g/t Au in drill hole SGRD962 from 236m including
  • 4m at 32.7 g/t Au from 26m
• 9 m at 4.8 g/t Au in drill hole SGRD955 from 246m including
  • 1m at 16.7 g/t Au from 252m

Click here to view the full listing of drill results from the recent drilling program at Koula. Note: all results are reported as down-hole intervals which represent approximately 60% of true width.

Séguéla Regional Reconnaissance

Scout RC drilling has recently been completed at Winy and will progressively work across the high priority Séguéla prospects. This testwork program is following up on the mapping and reconnaissance sampling at Séguéla which continues to emphasize the regional prospectivity of the property package with several prospects identified where rock chip samples recorded several instances of visible gold. In addition to ongoing detailed field mapping, auger drilling is underway targeting the southern extensions of key favourable structural corridors that host the Antenna, Ancien-Boulder-Agouti, and Elephant-Winy mineralization. Follow up scout RC drilling is planned in early 2021 across these prospects.

Mineral Resource Modeling and Estimation Assumptions

The Mineral Resources reported herein have been estimated using a geostatistical block modelling approach informed from gold assay data collected from RC and DD drill holes. Technical work for the construction of the Séguéla Mineral Resource, including wireframing, statistical and geostatistical analysis, and Mineral Resource estimation was conducted, and/or supervised by Mr. Hans Andersen of Roxgold and reviewed by Mr. Paul Weedon, VP of Exploration for Roxgold.

The Mineral Resource estimate incorporates all drilling data (RC and DD), comprised of 820 drill holes for a total of 110,936 metres targeting the Antenna, Agouti, Ancien, Boulder and Koula deposits.

The Koula, Ancien, Agouti and Boulder Mineral Resource models were developed using Leapfrog Geo and Micromine software. Antenna’s Mineral Resource model was developed using Geovia’s Surpac software. All gold assays from drill holes were composited to 1.0-meter intervals within the mineralised wireframes at Antenna, Agouti, Ancien, Boulder and Koula. Top-cuts were applied to individual domains based on the analysis of gold grade outliers within the statistical data populations and ranged between 1.5 g/t to 80.0 g/t Au.

The Mineral Resource model gold grades were estimated using a combination of Ordinary Kriging and Inverse Distance methods using a multiple pass approach to inform the resource model. The grade estimates are validated visually by sectional comparison and through statistical approaches that encompass traditional validation methods, such as Swath plots comparing composite and block model values for each deposit. Geostatistical exploratory data analysis (“EDA”), variogram modelling and Mineral Resource model validation was conducted using Snowden Supervisor software.

Density values were assigned to the Mineral Resource models based on ascribed oxidisation state and lithological unit, with mineralisation being assigned the density of its predominant host. A density of 1.2 to 1.8 t/m³ was assigned to transported and alluvial sediments, with a range of 1.8 to 2.2 t/m³ assigned to the oxidised weathered profile and a range of 2.67 to 3.20 t/m³ assigned to fresh rock lithologies.

Resource models and drill hole data at the Séguéla gold project utilise the WGS84 (Zone 29N) coordinate system.

The Mineral Resource models are classified into Measured, Indicated and Inferred categories based on analysis of the following criteria; number of samples informing the estimate, sample spacing, average sample distance, kriging efficiency and slope of regression outputs, drill hole and sample QAQC thresholds and geological confidence in modelled interpretations, grade continuity and level of geological understanding at each deposit.

The Séguéla Mineral Resources were reported constrained by preliminary pit optimisations generated in Micromine to satisfy the definition of Mineral Resources having reasonable prospects for eventual economic extraction (“RPEEE”), and are based on the following parameters:

• Gold price of US$1,700/oz
• Mining recovery of 90% and mining dilution of 10%
• Processing recovery of 94.5%
• Overall slope angle of 52° to 58° for Antenna, 54° for Agouti, 55° for Ancien, 55° for Koula and 57° for Boulder
• Mining costs of US$1.97 per tonne for Antenna and US$2.28 per tonne for the satellite deposits
• Total processing costs (including G&A) of US$21.64 per tonne
• Total selling costs (includes state and third-party royalties) of US$121.60/oz

Upcoming Catalysts

Qualified Persons

Paul Weedon, a Member of the Australian Institute of Geoscientists MAIG (#6001), Vice-President, Exploration for Roxgold Inc., a Qualified Person within the meaning of National Instrument 43-101, has reviewed, verified, and approved the scientific and technical disclosure contained in this news release. This includes the QA/QC, sampling, analytical and test data underlying this information. No limits were imposed on Mr. Weedon’s verification process.

The scientific and technical information contained in this document relating to the Séguéla Mineral Resource is based on, and fairly represents, information compiled by Mr. Hans Andersen. Mr. Hans Andersen, MAIG (#5746), is a Member of the Australian Institute of Geoscientists. Mr. Andersen is a full-time employee of Roxgold Inc. and is not “independent” within the meaning of NI 43-101 Standards of Disclosure for Mineral Projects. Mr. Andersen has sufficient experience which is relevant to the style of mineralisation and type of deposit under consideration and to the activity which he is undertaking to qualify as a “Qualified Person” under NI 43-101. Mr. Andersen has consented to and approved the inclusion in this document of the matters based on his compiled information in the form and context in which it appears in this document.

Roxgold’s disclosure of Mineral Reserve and Mineral Resource information is governed by NI 43-101 and under the guidelines set out in the Canadian Institute of Mining, Metallurgy and Petroleum (“CIM”) Standards on Mineral Resources and Mineral Reserves, adopted by the CIM Council, as may be amended from time to time by the CIM. There can be no assurance that those portions of Mineral Resources that are not Mineral Reserves will ultimately be converted into Mineral Reserves.

For more information on the projects you can also refer to the Company’s technical reports entitled “Technical Report for the Yaramoko Gold Mine, Burkina Faso” dated December 20, 2017 and “NI 43-101 Technical Report, Séguéla Project, Preliminary Economic Assessment, Worodougou Region, Cote d’Ivoire” dated April 14, 2020, each available on the Company’s website at www.roxgold.com and SEDAR at www.sedar.com.

Quality Assurance/Quality Control

All drilling data completed by Roxgold utilized the following procedures and methodologies. All drilling was carried out under the supervision of Roxgold personnel.

RC drilling used a 5.25-inch face sampling pneumatic hammer with samples collected into 60 litre plastic bags. Samples were kept dry by maintaining enough air pressure to exclude groundwater inflow. If water ingress exceeded the air pressure, RC drilling was stopped, and drilling converted to diamond core tails. Aircore (“AC”) drilling was collected in one metre intervals and sampled in a similar fashion to RC methods. Once collected, RC and AC samples were riffle split through a three-tier splitter to yield a 12.5% representative sample for submission to the analytical laboratory. The residual 87.5% sample were stored at the drill site until assay results were received and validated. Coarse reject samples for all mineralized samples corresponding to significant intervals are retained and stored on-site at the Company controlled core yard.

Diamond drill holes were drilled with HQ sized diamond drill bits. The core was logged, marked up for sampling using standard lengths of one metre. Samples were then cut into equal halves using a diamond saw. One half of the core was left in the original core box and stored in a secure location at the Company core yard at Séguéla. The other half was sampled, catalogued and placed into sealed bags and securely stored at the site until shipment.

All Séguéla RC, AC and DD core samples were shipped to ALS Laboratories preparation laboratory in Yamoussoukro for preparation. ALS Laboratories is independent of Roxgold. Samples were dried and crushed by the laboratory and a 250-gram split prepared from the coarse crushed material, prior to pulverization and preparation of a 200g sample. Samples are then shipped via commercial courier to ALS’s analytical facility in Ouagadougou, Burkina Faso where routine gold analysis using a 50-gram charge and fire assay with an atomic absorption finish was completed. Quality control procedures included the systematic insertion of blanks, duplicates and sample standards into the sample stream. In addition, the laboratory inserted its own quality control samples.