Sediment bioassays are becoming widely recognized as effective tools to determine the biological significance of the contamination found in coastal sediments. This paper describes the application of a new sediment toxicity test with Ampelisca abdita, a marine amphipod, to sediments from three stations at each of four sites in San Francisco Bay and one reference site in Tomales Bay. The tests were conducted under static and flow-through conditions, and all treatments and controls were maintained in quintuplicate (20 organisms per replicate). The exposure chambers were monitored daily for mortalities and emerged amphipods. Results indicated that the range of mortality within each station was generally less than 10%. Statistical analyses indicated highly significant site differences (p < 0.001) for both mortality and emergence. San Pablo Bay and Tomales Bay sediments were, as expected, the least toxic, with mean ten-day mortalities of 8.7 and 11.0%, respectively. Yerba Buena and Vallejo sediments were slightly, but significantly, more toxic at 15.0 and 14.7%, respectively; and the Oakland Harbor sediments were the most toxic, with 25.7% mortality. The data indicate that this sediment toxicity test can differentiate small differences in sediment quality. The emergence data were surprisingly consistent and showed the same response pattern, indicating that this sublethal response may be equal in sensitivity to survival. Static tests appeared slightly more sensitive than tests conducted under flow-through conditions.