Sparus aurata biomarkers were used to rank sediments from the Sado River estuary (Portugal) according to their toxicity. Initially, the activities of liver ethoxyresorufin-O-deethylase, liver and gill glutathione S-transferases, muscle lactate dehydrogenase, and brain acetylcholinesterase were tested in a laboratory bioassay with the reference compound benzo[a]pyrene. Enzymatic activities were determined in different tissues of fish exposed for 48, 96, or 240 h to three concentrations of benzo[a]pyrene (25, 50, and 100 mu g/L). Induction of liver ethoxyresorufin-O-deethylase was observed at all the exposure periods and concentrations, suggesting a continuous response of this system to toxicant exposure. Induction of liver glutathione S-transferases activity was only observed after 240 h of exposure, whereas gill glutathione S-transferases activity was significantly inhibited at all the exposure periods, suggesting a direct or indirect effect of the toxicant on these enzymes. Inhibition of lactate dehydrogenases activity was only observed after 96 h of exposure to 25 mu g/L of benzo[a]pyrene. No significant effects were observed on acetylcholinesterase activity, suggesting that cholinergic function of S. aurata is not affected by benzo[a]pyrene. In a second phase, fish were exposed for 240 It to sediments collected at five sites of the Sado River estuary, and the same biomarkers were analyzed. For all the enzymes assayed, significant differences among sites were found. In this study, the battery of biomarkers used allowed to discrimination among sites with different types of contamination, levels of contamination, or both, after multivariate data analysis. Discrimination of sites was similar to the ranking provided by a more complex and parallel study (including chemical analysis of sediments, macrobenthic community analysis, amphipod mortality toxicity tests, and sea urchin abnormality embryo assays), suggesting its suitability to evaluate the toxicity of estuarine sediments.