A battery of in vitro short-term tests revealing different genetic end-points was set up in order to study surface-water genotoxicity after disinfection with different biocides: sodium hypochlorite (NaClO), chlorine dioxide (ClO(2)) and peracetic acid (PAA). The surface water both before and after disinfection was concentrated by adsorption on C(18) silica cartridges and the concentrates containing non-volatile organics were divided into different portions for chemical analyses and biological assays. The following in vitro tests were conducted on the water concentrates dissolved in DMSO: the Salmonella mutagenicity assay with S. typhimurium strains TA98 and TA100; the SOS Chromotest with Escherichia coli, the Microtox and Mutatox assays with Vibrio fischeri; and gene conversion, point mutation and mitochondrial DNA mutability assays with D7 diploid Saccharomices cerevisiae strain. The results show that the SOS Chromotest and the yeast assays are highly sensitive in detecting genotoxicity. The surface-water extracts were very often toxic to most of the test organisms considered, partially masking their potential mutagenic activity. Therefore, the assays with E. coli and with S. cerevisiae are more likely to show a mutagenic effect because these organisms are generally less sensitive to most toxic compounds. Among the tested disinfectants, NaClO and ClO(2) increased water genotoxicity, whereas PAA was able to slightly reduce raw water activity. However, because the organic compounds in the lake water varied with the season of the year, the disinfection processes, at times, both increased and decreased the raw water activity.