Thermochemical sulfate reduction (TSR) is one of the most important organic-inorganic reactions in petroleum reservoirs which significantly affects production and processing risks. The initiation (non-catalyzed) stage of TSR is critical in constructing a reliable kinetic model to predict this reaction. There is a large gap in our understanding of the mechanisms involved in this stage. In the present study, we used hydrous pyrolysis experiments with Na2SO4 and 1-dodecene as model compounds at pH= 1, and temperatures of 250 °C and 300 °C to simulate this early TSR stage. We measured and calculated the sulfur isotopic fractionations of sulfate, H2S and individual organic sulfur compounds (OSC). A thermodynamic model was used to determine the concentrations of the different sulfate species (H2SO4, HSO4¯, SO42-) at the experimental conditions. We then used ab initio calculations to determine possible radical and non-radical reaction mechanisms to explain the experimental results. During all the experiments, sulfate was consumed gradually while H2S and OSC were produced. The residual sulfate was 34S enriched up to 10.7‰ relative to the initial sulfate as pyrolysis time advanced. The produced H2S and OSC were typically about 21‰ 34S depleted relative to co-existing sulfate. The formation of OSC, such as C24 sulfides and C12 thiophenes, during the experiments and their similar δ34S values, suggest a common precursor and that radical mechanisms are involved in their formation. Calculated activation energies and free energies (Ea and ∆G‡, respectively) for the radical TSR mechanisms (20.6–34.1 and 32.3–44.4 kcal mol−1, respectively) yielded lower values than the non-radical TSR mechanisms (53.5–68.5 and 59.7–78.1 kcal mol−1, respectively). However, the calculated kinetic fractionation factors for the non-radical mechanism correlated better with the 11–12‰ experimental fractionation factor of current and previous studies. This similarity between the experimental and computational results indicates that these non-radical rather than radical mechanisms were likely involved in the reduction of sulfate in our experiments. Despite the lower activation energies, the radical TSR mechanisms are not dominant in our experimental system, probably due to competing reactions of the active radical species (alkyl and allyl radicals). Nevertheless, under natural conditions, the significantly lower amount of alkenes is expected to reduce the competition on the active alkyl and allyl radicals for alkylation, making these radical mechanisms competing pathways for the initiation stage of TSR.