Sulfur isotope fractionation during dissimilatory sulfate reduction has been conceptually described by the widely accepted Rees model as related to the stepwise reduction of sulfate to sulfide within the cells of bacteria. The magnitude of isotope fractionation is determined by the interplay between different reduction steps in a chain of reactions. Here we present a revision of Rees’ model for bacterial sulfate reduction that includes revised fractionation factors for the sulfite-sulfide step and incorporates new forward and reverse steps in the reduction of sulfite to sulfide, as well as exchange of sulfide between the cell and ambient water. With this model we show that in contrast to the Rees model, isotope fractionations well in excess of −46‰ are possible. Therefore, some of the large sulfur isotope fractionations observed in nature can be explained without the need of alternate pathways involving the oxidative sulfur cycle. We use this model to predict that large fractionations should occur under hypersulfidic conditions and where electron acceptor concentrations are limiting.