The antiemetic activity of two different high doses and schedules of metoclopramide in dacarbazine-treated cancer patients was compared in a double-blind crossover study. Regimen A consisted of metoclopramide [2 mg/kg x 4 intravenously (i.v.)] plus methylprednisolone (250 mg x 2 i.v.) plus diphenhydramine (50 mg x 2 i.v.). Regimen B consisted of metoclopramide (3 mg/kg x 2 i.v.) plus dexamethasone (20 mg i.v.) and diphenhydramine (50 mg i.v.). Both treatments were administered for the first 2 days of 5-day dacarbazine chemotherapy. Thirty-two patients (13 men and 19 women) affected by melanoma and sarcoma were entered in the study. Complete protection against nausea and vomiting for the first 2 days of chemotherapy in both antiemetic regimens was not significantly different. Patient preference and tolerance of the two antiemetic treatments were similar. Regimen B, employing a lower dosage of metoclopramide and steroids and using a more simple schedule of administration should be the preferred treatment.