To evaluate the genotoxic risk that contaminated sediment could constitute for benthic organisms, three contaminated (VA, VC and VN) and one uncontaminated (RN) sediment samples were collected in the Berre lagoon (France). Potentially bioavailable contaminants in sediments were obtained using sediment extraction with synthetic seawater adjusted to pH 4 or pH 6, simulating the range of pH prevailing in the digestive tract of benthic organisms. The genotoxic activities of these extracts were evaluated by three short-term bioassays: the Salmonella mutagenicity test using the Salmonella typhimurium strain TA102, the alkaline comet assay and the micronucleus assay on the Chinese Hamster Ovary cells CHO-K1. Results of the Salmonella mutagenicity assay detected a mutagenic response for RN extract at pH 6, and for VA extract at pH 4. Results of the comet and micronucleus assays detected low genotoxic/clastogenic activities for VA and VC extracts at pH 6 and higher activities for RN. VA and VC extracts at pH 4. To identify if metals (Al, Fe, Mn, As, Cd, Co, Cr, Cu, Hg, Ni, Pb and Zn) were involved in these genotoxic activities, their concentrations were determined in the extracts, and their speciation was assessed by thermodynamic calculations. Results showed that extracts from sites VA. VC and VN generally presented the highest trace metal contents for both extractants, while the site RN presented lower trace metal contents but the highest Fe and Mn contents. Thermodynamic calculations indicated that Fe, Mn, As and in a lower extend Co, Ni and Zn were mainly present under free forms in extracts, and were consequently, more likely able to induce a genotoxic effect. Results globally showed no correspondence between free metal contents and genotoxic activities. They suggested that these positive results could be due to uncharacterized compounds, acting as direct genotoxic agents or enhancing the genotoxic properties of analyzed metals. (C) 2012 Elsevier Ltd. All rights reserved.