Bacterial sulfate reduction activity (SRA) was measured in surface sediments and slurries from three sites in the Great Salt Lake (Utah, USA) using radiolabeled 35S-sulfate. High rates of sulfate reduction (363 +/- 103 and 6,131 +/- 835 nmol cm-3 d-1) were measured at two sites in the moderately hypersaline southern arm of the lake, whereas significantly lower rates (32 +/- 9 nmol cm-3 d-1) were measured in the extremely hypersaline northern arm. Bacterial sulfate reduction was strongly affected by salinity and showed an optimum around 5-6% NaCl in the southern arm and an optimum of around 12% NaCl in the more hypersaline northern arm of the lake. High densities of sulfate-reducing bacteria (SRB) ranging from 2.2 x 107 to 6.7 x 108 cells cm-3 were determined by a newly developed tracer MPN-technique (T-MPN) employing sediment media and 35S-sulfate. Calculation of specific sulfate reduction rates yielded values comparable to those obtained in pure cultures of SRB. However, when using a conventional MPN technique with synthetic media containing high amounts of Fe(II), the numbers of SRB were underestimated by 1-4 orders of magnitude as compared to the T-MPN method. Our results suggest that high densities of slightly to moderately halophilic and extremely halotolerant SRB are responsible for the high rates of sulfate reduction measured in Great Salt Lake sediments.