We used 86Rb+ (K+ analogue) to study potassium influx during the interaction of highly specific 5-HT3-receptor antagonists, ondansetron and granisetron, with the effects of the anticancer drug, estramustine phosphate, on P31 mesothelioma cells. Estramustine phosphate (80 mg/l, 142 micromol/l) for 120 min. reduced 86Rb+ influx by 18.7%. The reduction was inhibited by ondansetron (0.1 micromol/l), but augmented by granisetron (0.1 micromol/l). Serotonin (1.0 micromol/l) antagonized ondansetron inhibition and restored granisetron-augmented reduction of estramustine phosphate-induced 86Rb+ influx to the level of the drug itself. Estramustine phosphate inhibited cellular Na+, K+, 2Cl- -cotransport activity whereas Na+, K+, ATPase activity was unaffected. Ondansetron blockade of estramustine phosphate-induced reduction of 86Rb+ influx was due to increased Na+, K+, ATPase and Na+, K+, 2Cl- -cotransport whereas augmentation of estramustine phosphate-induced reduction of 86Rb+ influx by granisetron, or combination of 5-HT3 receptor antagonists with serotonin was due mainly to inhibition of cellular Na+, K+, ATPase activity Thus, ondansetron possesses a distinct ability to reverse K+ influx of tumour cells exposed to estramustine phosphate whereas granisetron does not, due to different effect on cellular Na+, K+, ATPase and Na+, K+, 2Cl- -cotransport activity. Highly 5-HT3 receptor-specific antiemetic agents may have different effects on ion transport of tumour cells during treatment with cytotoxic drugs.