The Comet assay has been used widely in genetic toxicology, radiation biology and medical and environmental research. This assay detects single-strand breaks and alkali-labile sites in DNA and DNA degradation due to necrosis or apoptosis. It may also be modified to detect DNA cross-linking. Although a considerable number of chemicals have been tested in the assay there are many aspects of validation to be considered before the method could be considered to provide definitive evidence of genotoxic potential. For example, very few non-genotoxins have been tested to assess specificity of the Comet assay and there has been only one reported study which investigated whether the in vitro Comet assay is prone to false positive responses due to cytotoxicity. We have investigated the response of the alkaline Comet assay in TK6 human lymphoblastoid cells to cytotoxic damage and genotoxic damage. Several compounds which are toxic by different mechanisms were tested in the assay. Cycloheximide and trypsin gave a negative comet response at a highest dose of 5 mg/ml and no toxicity was observed. Sodium lauryl sulphate and potassium cyanide produced a significant increase in DNA migration at cell survival levels of < or = 75%. The distribution of damaged cells indicated that cells at various stages of necrotic cell death were present. Hydrogen peroxide, 4-nitroquinoline oxide, 9-aminoacridine, ethyl methanesulphonate, N-nitroso-N-ethylurea and glyoxal gave a positive comet response. Mitomycin C was negative at survival levels of approximately 70%. These results indicate that the maximum concentration of test substance tested should produce viabilities > 75% in order to avoid false positive responses due to cytotoxicity. The assay was able to detect DNA damage induced by an alkylating agent, an intercalating agent and oxidative damage. The cross-linking agent mitomycin C was not detected if a cut-off point of 75% viability is used as the criterion of a positive response.