The genotoxicity of river water and sediment including interstitial water was evaluated by microscreen phage-induction and Salmonella/microsome assays. Different processes used to fractionate the sediment sample were compared using solvents with different polarities. The results obtained for mutagenic activity using the Salmonella/microsome test were negative in the water and interstitial water samples analysed using the direct concentration method. The responses in the microscreen phage-induction assay showed the presence of genotoxic or indicative genotoxic activity for at least one water sample of each site analysed using the same concentration method. Similar results were obtained for interstitial water samples, i.e. absence of mutagenic activity in the Salmonella/microsome test and presence of genotoxic activity in the microscreen phage-induction assay. Metal contamination, as evidenced by the concentrations in stream sediments, may also help explain some of these genotoxic results. Stream sediment organic extracts showed frameshift mutagenic activity in the ether extract detected by Salmonella/microsome assay. The concentrates evaluated by microscreen phage-induction assay identified the action of organic compounds in the non-polar, medium polar and polar fractions. Thus, the microscreen phage-induction assay has proven to be a more appropriate methodology than the Salmonella/microsome test to analyse multiple pollutants in this ecosystem where both organic compounds and heavy metals are present.