Relative cytotoxicity and genotoxicity of a widely used antibiotic, trimethoprim (TRIMP) was evaluated under in vitro conditions using rainbow trout gonad-2 (RTG-2) and Chinese hamster ovary-K1 (CHO-K1) cells. Whilst cytotoxicity was determined using neutral red retention (NRR) assay, the genotoxicity was determined using single cell gel electrophoresis or the Comet assay and cytokinesis-block micronucleus (CBMN) assay. For NRR assay, concentration-dependent cytotoxic effect was observed for both the cell lines (estimated EC(50) values: 671.82 ± 21.78 and 611.6 ± 20.4 μg ml(-1) for RTG-2 and CHO-K1 cells, respectively). There was no statistically significant difference between the two cell lines for this assay. For the Comet assay, standard 6 h exposure to TRIMP did not show any positive response for any of the cell types used. However, 48 h exposure to RTG-2 cells showed a concentration-dependent induction of DNA damage (r = 0.86). The highest concentration of TRIMP used (i.e. 100 μg ml(-1)) showed relatively higher DNA damage, compared to ethyl methane sulfonate (EMS; 1 μg ml(-1) or 8 mM), a reference genotoxic agent, used concurrently. In contrast, 24 h exposure time for CHO-K1 cells did not show any concentration-dependent increase for this assay. For MN assay, a significant correlation was found between the MN induction and TRIMP concentration for both the cell lines (RTG-2: r = 0.68; CHO-K1: r = 0.79), although only the highest concentration used showed a significant increase for binucleated (BN) cell with micronuclei (BNMN). The study suggests that whilst the cells of different origin could exhibit similar cytotoxicity, they could display differential genotoxic effects. Furthermore, genotoxic effects of TRIMP are primarily exposure period dependent phenomena and, in addition to inhibiting the action of dihydrofolate reductase, oxidative stress could also contribute for the observed toxic effects, fish cells in general being more sensitive for genotoxic effects.