Cisplatin (8 mg/kg) was given intravenously to guinea pigs either as a 15 s bolus injection (25 animals) or as a 1 h infusion (28 animals). To determine the influence of the mode of cisplatin administration and pharmacokinetics on the ototoxic side-effect, the concentrations of cisplatin and the biotransformation product monoaquated cisplatin were determined in blood ultrafiltrate using liquid chromatography with post-column derivatization. Ototoxic effect was evaluated as difference in pre- and 96 h post-exposure auditory brainstem response (ABR) threshold. The cisplatin peak concentration was considerably higher, 19.2+/-2.4 microg/ml, in the bolus injection group than in the infusion group, 6.7+/-0.5 microg/ml (mean+/-S.E.M.). The area under the blood ultrafiltrate concentration time curve (AUC) for cisplatin was slightly greater in the infusion group, 442+/-26 microg/ml/min, than in the bolus injection group, 340+/-5 microg/ml/min. For monoaqua cisplatin, the AUC was not different between the groups (bolus injection: 30.8+/-1. 5 microg/ml/min, infusion: 34.1+/-3.3 microg/ml/min). A significant ototoxic effect was observed in both groups at 20 and 12.5 kHz, but there was no difference between the groups in the extent of threshold shift. The interindividual variability in susceptibility to ABR threshold shift was far greater than the variability in pharmacokinetics, suggesting that other factors are more important in determining the degree of hearing loss.