The antiemetic efficacy of haloperidol was studied using data from 15 published (1962-1988) and 8 unpublished randomized trials; 1,397 adults received haloperidol, and 1,071 were controls. Settings were postoperative nausea or vomiting (1,994 patients), gastroenterology (261), chemotherapy (189), and radiation therapy (24). The relative benefit to prevent postoperative nausea or vomiting during 24 h with 0.5-4 mg haloperidol compared with placebo was 1.26-1.51 (number needed to treat, 3.2-5.1), without evidence of dose responsiveness; 0.25 mg was not antiemetic. With 1 mg haloperidol, the relative benefit to stop postoperative nausea or vomiting during 2-4 h compared with placebo was 1.53 (95% confidence interval, 1.17-2.00; number needed to treat, 6); with 2 mg, the relative benefit was 1.73 (1.11-2.68; number needed to treat, 4). In gastroenterology, 2 mg haloperidol was more effective than 1 mg. For chemotherapy and radiation therapy, no conclusions could be drawn. With 4 mg, one patient had extrapyramidal symptoms. With 5 mg, sedation was increased, with a relative risk of 2.09 (95% confidence interval, 1.73-2.52; number needed to treat, 4.4). There were no reports on cardiac toxicity. Postoperatively and in gastroenterology, haloperidol is antiemetic, with minimal toxicity. For other clinical settings and for children, valid data are unavailable.