A series of consecutive trials were undertaken to determine whether higher doses of intravenous metoclopramide and combinations of metoclopramide, dexamethasone, and diphenhydramine would improve antiemetic control or decrease treatment-related side effects in patients receiving cisplatin at 120 mg/ m2. Metoclopramide and dexamethasone were studied because of their proven efficacy as single agents and their differing mechanisms of action and side effects. Diphenhydramine was used because of its possible antiemetic properties and its ability to control acute dystonic reactions. Two hundred fifty-five patients who had never received chemotherapy or antiemetics were observed in the hospital for the 24 hours following cisplatin administration. The addition of dexamethasone or dexamethasone plus diphenhydramine to intravenous metoclopramide 2 mg/kg produced both improved antiemetic control and a decrease in treatment-associated diarrhea (P = 0.002). The use of metoclopramide alone at a dose of 3 mg/kg for only two doses appeared as effective as 2 mg/kg for five doses. When dexamethasone and diphenhydramine were given with metoclopramide 3 mg/kg for two intravenous dosages, 81% of patients experienced no emesis and 93% had two or fewer vomiting episodes. The antiemetic results of this 2-hour "short-course" regimen were superior to metoclopramide 2 mg/kg, with (P = 0.002) or without (P = 0.0001) dexamethasone and diphenhydramine. It was concluded that combinations of metoclopramide plus dexamethasone plus diphenhydramine improve antiemetic control, facilitate the usage of higher doses of metoclopramide, and decrease the incidence of treatment-related side effects.