Sediment quality was assessed in San Francisco Bay, California, USA, using a two-tiered approach in which 111 sites were initially screened for sediment toxicity. Sites exhibiting toxicity were then resampled and analyzed for chemical contamination, recurrent toxicity, and, in some cases, benthic community impacts. Resulting data were compared with newly derived threshold values for each of the metrics in a triad-based weight-of-evidence evaluation. Sediment toxicity test results were compared with tolerance limits derived from reference site data, benthic community data were compared with threshold values for a relative benthic index based on the presence and abundance of pollution-tolerant and -sensitive taxa, and concentrations of chemicals and chemical mixtures were compared with sediment quality guideline-based thresholds. A total of 57 sites exceeded threshold values for at least one metric, and each site was categorized based on triad inferences. Nine sites were found to exhibit recurrent sediment toxicity associated with elevated contaminant concentrations, conditions that met program criteria for regulatory attention. Benthic community impacts were also observed at three of these sites, providing triad evidence of pollution-induced degradation. Multi- and univariate correlations indicated that chemical mixtures, heavy metals, chlordanes, and other organic compounds were associated with measured biological impacts in the Bay. Toxicity identification evaluations indicated that metals were responsible for pore-water toxicity to sea urchin larvae at two sites. Gradient studies indicated that the toxicity tests and benthic community metrics employed in the study predictably tracked concentrations of chemical mixtures in Bay sediments.