Selenium stable isotope ratios can serve as indicators of Se sources and reduction of Se oxyanions, much as sulfur and nitrogen isotope ratios do in sulfur and nitrogen biogeochemical studies. A new analytical method, which allows precise Se isotope ratio measurements on 500 ng of Se, greatly enhances analysis of environmental samples. This paper presents the first environmental study to use Se stable isotopes. 80Se/76Se ratios, relative to a provisional standard, were measured in water, oil refinery wastewater, total sediment digest, and sediment extracts from the Carquinez area in the San Francisco Bay Estuary, Se isotope ratios in total sediment Se and in extracts designed to recover Se0 are slightly (about 2???) enriched in the lighter isotope relative to local bay water Se. This difference is smaller than the isotopic fractionations expected upon reduction of Se(VI) or Se(IV) to Se(0) and suggests that reduction of soluble selenium from the overlying waters is not the dominant process by which Se is incorporated into the sediments. Consistent isotopic differences between riverine and refinery inputs were not observed, and thus tracing of refinery inputs with Se isotopes is not possible in this system.Selenium stable isotope ratios can serve as indicators of Se sources and reduction of Se oxyanions, much as sulfur and nitrogen isotope ratios do in sulfur and nitrogen biogeochemical studies. A new analytical method, which allows precise Se isotope ratio measurements on 500 ng of Se, greatly enhances analysis of environmental samples. This paper presents the first environmental study to use Se stable isotopes. 80Se/76Se ratios, relative to a provisional standard, were measured in water, oil refinery wastewater, total sediment digests, and sediment extracts from the Carquinez area in the San Francisco Bay Estuary. Se isotope ratios in total sediment Se and in extracts designed to recover Se0 are slightly (about 2 per mil) enriched in the lighter isotope relative to local bay water Se. This difference is smaller than the isotopic fractionations expected upon reduction of Se(VI) or Se(IV) to Se(0) and suggests that reduction of soluble selenium from the overlying waters is not the dominant process by which Se is incorporated into the sediments. Consistent isotopic differences between riverine and refinery inputs were not observed, and thus tracing of refinery inputs with Se isotopes is not possible in this system.