We designed a double-blind randomized crossover study to compare the antiemetic activity and toxicity of high-dose intravenous alizapride (3.5 mg/kg) versus high-dose intravenous metoclopramide (1 mg/kg) both combined with intravenous methylprednisolone in 40 untreated cancer patients submitted to cisplatin chemotherapy alone. The mean number of vomiting episodes (3.6 vs. 1.0), length of vomiting (209.8 vs. 80.9 min), and rate of complete prevention of vomiting (25% vs. 68.4%) at first cycle were in favor of metoclopramide, with a statistically significant difference. This difference was not statistically significant in the second cycle after crossover. Toxicity of both treatments was mild and diarrhea was more frequent in the alizapride-treated patients. Preference expressed by the patients was also in favor of metoclopramide. We conclude that alizapride offers less antiemetic protection than metoclopramide in patients receiving cisplatin chemotherapy. Further use of alizapride in such patients is not recommended.