We compared the antiemetic efficacy of metoclopramide in a bolus low-dose infusion schedule to that of metoclopramide given in a conventional high-dose bolus schedule in a randomized crossover trial. Thirty-two treatment courses in 16 patients receiving cisplatin chemotherapy were evaluable. The metoclopramide regimen was either 2 mg/kg i.v. bolus, then 20 mg/h by infusion for 4 h, or 2 mg/kg i.v. bolus every 2 h for three doses. Dexamethasone 20 mg i.v. and diphenhydramine 50 mg i.v. were also given. Antiemetic efficacy was assessed by a questionnaire. There were no differences in antiemetic efficacy between the metoclopramide regimens. With either program, 75% of patients were emesis-free, 13% had mild symptoms, and 13% had moderate symptoms (greater than two emetic episodes). The infusion metoclopramide regimen was 30% less expensive than the bolus schedule in our pharmacy. Thus, we recommend low-dose metoclopramide infusion as a less expensive, equally effective alternative to high-dose bolus regimens for antiemetic treatment.