We compared the antiemetic efficacy of ondansentron versus ondansentron and corticosteroids in cisplatin-induced emesis. None of our patients had received prior chemotherapy. All patients received chemotherapy including cisplatin 100 mg/m2. Forty patients received ondansentron alone (Group A) and 40 the combination of ondansentron and methylprednisolone (Group B). Ondansentron was given at a dose of 8 mg in 100 mL N/S over 10 min by intravenous infusion. The initial dose was administered before the cisplatin and was followed by 8 mg orally in the afternoon and before sleeping the first day of chemotherapy. During the next 2 days, the patients received 8 mg orally 3 times daily. Methylprednisolone was given as an intravenous bolus of 40 mg before chemotherapy and then together with each dose of ondansentron at a dose of 16 mg orally. Group A had significantly longer duration of nausea after chemotherapy than group B (117 +/- 111 min, 62 +/- 71 min, P < 0.013). The response on emesis was also improved in group B, especially the day of chemotherapy [treatment failure: group A: 13 patients (30%) versus group B: 5 patients (11.6%), P < 0.03] and the next day [complete response: group A: 17 patients (39.5%) versus group B: 30 patients (69.7%), P < 0.005]. Patients in group B presented more sedative effects (P < 0.001) and better appetite (P < 0.02) than patients in group A. There were no other significant differences in side effects (activity, headache, constipation, etc). We conclude that corticosteroids improve the antiemetic efficacy of ondansentron in cisplatin-induced chemotherapy, and should be included in antiemetic regimens.