In 29 patients (48 courses of chemotherapy), a randomized double-blind crossover study was undertaken to establish the antiemetic effectiveness of a combination of metoclopramide at a moderately high dose (5 mg/kg body weight) with 1 g methylprednisolone (treatment A) as compared with metoclopramide monotherapy (5 mg/kg body weight; treatment B). With cisplatin doses of 60 mg/m2, a good antiemetic effect (0-1 vomiting episodes in 24 h) was observed in 9 out of 10 patients under treatment A as well as under treatment B. With a cisplatin dose of 120 mg/m2, treatment A produced a good antiemetic effect in 9 out of 11 patients (82%), while treatment B produced such an effect in only 3 out of 11 patients (27%; p less than 0.5). The results obtained permit the following conclusions to be made: At doses of cisplatin of up to 60 mg/m2, the antiemetic effect produced by metoclopramide (5 mg/kg body weight) given alone is adequate. At doses of cisplatin of 120 mg/m2 and more, high-dose methylprednisolone has a good additive antiemetic effect vis-à-vis metoclopramide (5 mg/kg body weight). The antiemetic effect of this combination corresponds approximately to the antiemetic effect produced with high-dose metoclopramide (10 mg/kg body weight) administered alone.