Investigations on geochemical processes involved in H 2 generation in natural reservoirs are important for the present energy transition from fossil fuels. Most studies related to native H 2 focus on the serpentinisation of ultramafic rocks, in which the source of H 2 is linked to oxidation of Fe 2þ-bearing minerals and a reduction of H 2 O. In this study, we investigate abiotic H 2 generation from a biotite-rich granite using an approach based on a study case of the Soultz-sous-Forêts geothermal site in France. Using a geochemical and reactive transport model and an existing database, we simulated the hydrothermal alteration of the granite at 130-200 °C with a redox potential from À 100 mV to À 300 mV. The simulations show that generation of native H 2 is possible in an open system by hydrothermal alteration of biotite as a source of Fe 2þ , which oxidises to Fe 3þ leading to precipitation of Fe 3þ minerals and reduction of H þ. The amount of H 2 produced depends on the type of Fe 2þ/3þ minerals which precipitate. The optimal conditions for H 2 generation are related to magnetite precipitation as in the serpentinisation process. We assessed that the upper bound for H 2 production for this site is 102 kt per km 3 of granite. This is in case of complete oxidation of all the ferrous iron contained in the biotite of the geothermal reservoir. In practice, only part of this potential may be attained owing to the limited current exploitation methods which are unlikely to result in a complete disequilibrium of the whole biotite. Future research needs to focus on how to reach most of the H 2 potential. Simulations with increasing CO 2 pressures suggest that CO 2 injection can stimulate the H 2 production. This study has implications for possible coupling of heat extraction and future exploitation with H 2 production.