Enzymatic reactions during dissimilatory sulfate reduction (DSR) are often treated as unidirectional with respect to dissolved sulfide. However, quantitative models describing kinetic sulfur isotope fractionations during DSR consider the individual enzymatic reactions as reversible (Rees, 1973). Brunner and Bernasconi (2005) extended this line of thought, and suggested that as long as cell external sulfide (CES) concentrations are high enough, CES may diffuse back across the cytoplasmic cell membrane and may subsequently be re-oxidized to sulfate. Here, we test this hypothesis by measuring the time evolution of the d 34 S-sulfate signal during DSR in closed system experiments under different levels of sulfide stress (0–20 mM and 0–40 mM total dissolved sulfide). Our results show that the measured d 34 S-sulfate signal is markedly different in the latter case and that the observed sulfate S-isotope time-evolution is incompatible with a Rayleigh type fractionation model. In contrast, our results are consistent with a sulfate reduction and fractionation model that allows for a cell internal oxidation of dissolved sulfide by a sulfate reducer. 2011 Elsevier Ltd. All rights reserved.