Twenty-six children aged 4 to 15 years who were to receive cancer chemotherapy were enrolled in a double-blind, randomized, crossover trial that compared the antiemetic efficacy of a four-drug regimen (the MBDL regimen: metoclopramide, 8 mg/kg; benztropine, 0.04 mg/kg; dexamethasone, 0.7 mg/kg; lorazepam, 0.1 mg/kg), given over 24 hours, with the efficacy of chlorpromazine, 3.3 mg/kg, given in four doses over 24 hours. The MBDL regimen was more effective than chlorpromazine in both objective and subjective measures of antiemetic control. Of 26 children, 23 (89%) had less vomiting on the MBDL regimen, and 20 (77%) of 26 patients or parents preferred this regimen (p less than 0.01). The MBDL regimen reduced the number of vomiting episodes by a mean of 4.0 (p less than 0.01) and reduced the duration of vomiting by a mean of 3.7 hours (p less than 0.01). A moderate level of sedation was documented at some stage in the 24-hour period of observation in 27% on the MBDL regimen and in 35% receiving chlorpromazine. Dystonia was seen in 1 (4%) of 26 children. We conclude that the MBDL regimen is safe in children and more effective than chlorpromazine.