PURPOSE: To compare the efficacy and safety of oral ondansetron with i.v. granisetron each given as a single dose prior to administration of highly emetogenic cisplatin chemotherapy. PATIENTS AND METHODS: Chemotherapy-naive patients with histologically confirmed malignancies were randomized to receive a single 24 mg ondansetron hydrochloride tablet plus a 50 ml i.v. infusion of normal saline, or a single 10 µg/kg (50 ml) i.v. infusion of granisetron plus a placebo tablet in this multicenter, double-blind, parallel-group trial. Study drug was administered 30 min prior to a single i.v. infusion of cisplatin (50-75 mg/m²), given over a period of </= 3 h. Concurrent administration of corticosteroids was not allowed. Efficacy measurements included the number of emetic episodes, need for rescue medication, and patient assessments of nausea and appetite. Complete response (CR) was defined as no emetic episodes, rescue, or withdrawal; major response was defined as one or two episodes. Safety was evaluated by monitoring adverse events and changes in laboratory parameters. RESULTS: A total of 371 patients entered the study and received study drug, of whom 184 received ondansetron and 187 received granisetron. For all parameters tested, a single 24 mg oral ondansetron tablet was at least as effective as i.v. granisetron. CR was achieved in 58% of ondansetron-treated patients and 51% of granisetron-treated patients (95% confidence interval on the difference: -4% to 17%). Subjective assessments revealed no difference with regard to complete control of nausea, appetite, or satisfaction with antiemetic therapy. Both drugs were well tolerated; the most common adverse event was headache. CONCLUSION: A single 24 mg oral dose of ondansetron is at least as safe and effective as a single i.v. infusion of 10 µg/kg of granisetron in preventing nausea and vomiting induced by highly emetogenic cisplatin chemotherapy.