We report here results obtained using sodium chloride and potassium chloride as model compounds to investigate the effects of high dose levels and osmolarity-mediated artefacts in mutagenicity assays using cultured mammalian cells. Three assay systems were used with different genetic end-points: (i) mutation to 6-thioguanine resistance in Chinese hamster V79 cells; (ii) induction of chromosome aberrations in Chinese hamster ovary cells; and (iii) induction of unscheduled DNA synthesis in HeLa S3 cells. In V79 cells we observed sporadic increases in mutation frequency after treatment with sodium chloride. Potassium chloride increased the mutation frequency in a narrow interval of dose levels. Chromosome aberrations were induced by both compounds at the highest concentrations tested, confirming previously published data; the effects of potassium chloride were more marked. The chromosome aberrations observed included both deletions and exchange figures. Neither compound induced UDS in our experiments. We performed the Ames test as a check for mutagenic contaminants and both compounds gave entirely negative results. Although the osmolarities of sodium chloride and potassium chloride solutions were similar, the effects of potassium chloride were always greater at equivalent concentrations; the nature of the solute, and not only the observed osmolarity, appears to influence the results obtained. Similarly, we have observed that dimethylsulphoxide causes a marked increase in the osmolarity of the culture medium without any obvious induction of chromosome aberrations. Although the presence of S9 had little effect on the osmolarity of the medium, differences in response were observed after treatments with or without S9. For the two model compounds it appears that an osmotic mechanism alone does not provide a sufficient explanation of the findings.