Many studies have revealed the presence of compounds with genotoxic activity in drinking water by means of short-term mutagenicity tests. In this study, the influence of the different steps of surface water treatment on the mutagenicity of drinking water was evaluated. Four different types of samples were collected: raw lake water, water after pre-disinfection with chlorine dioxide, water after filtration on granular activated carbon, and tap water. Water extracts underwent a bacterial toxicity test (Microtox test) and different in vitro genotoxicity tests: a test with Salmonella typhimurium strains, a Saccharomyces cerevisiae test, the SOS Chromotest with Escherichia coli and the Mutatox test with Vibrio fischeri. The Microtox test revealed high toxicity in the treated water samples. The disinfection steps increased the toxicity: the Mutatox test confirmed these results and the Salmonella/microsome test at the highest doses showed toxicity that could conceal mutagenicity. The SOS Chromotest was positive in all treated water samples without metabolic activation. In the test with S. cerevisiae both toxicity and genotoxicity generally increased during the water treatment steps, especially in cells without induction of cytochrome P450.