Phase II trials of ondansetron were undertaken to assess the ability of this agent to control nausea and vomiting caused by specific chemotherapeutic agents, to establish the optimal number of doses and the most appropriate schedule of administration, to see if control could be improved by the use of continuous infusion, and to ascertain if the degree of efficacy and safety of ondansetron would warrant further investigations. In each of six multiplebolus trials, ondansetron was given at 0.15 mg/kg to 0.18 mg/kg intravenously for three doses, beginning 30 minutes prior to cisplatin. No patient had received prior cancer chemotherapy. Overall, 48% of patients experienced no emesis, and 71% had zero to two emetic episodes after receiving cisplatin doses of 100 mg/m2 or greater. Comparable antiemetic control was seen with all schedules studied. Three additional trials assessed the effect of the number of doses of ondansetron on antiemetic effectiveness. A single dose gave complete protection in 25% of patients and three doses gave a 50% no-emesis rate. Complete control was not improved when six doses of ondansetron were given. Efficacy rates with multiple-bolus therapy and continuous infusion over 24 hours were similar. Side effects were mild and reversible in all trials, and there were no remarkable differences in adverse events among different schedules or numbers of doses. The complete and major control rates observed show that ondansetron is as effective as or more effective than metoclopramide in controlling cisplatin-induced emesis.