Nineteen patients receiving cancer chemotherapy were randomized in a double-blind fashion to receive either (a) dronabinol, 10 mg plus placebo q.i.d.; (b) prochlorperazine, 10 mg plus placebo q.i.d.; or (c) dronabinol plus prochlorperazine, each 10 mg q.i.d. There were six evaluable patients in each of the two single-agent groups and five in the combination group. The median duration and severity per episode of nausea was significantly greater in the group receiving prochlorperazine alone versus the other two groups. The median duration per episode of vomiting was also significantly greater in the prochlorperazine group than in the other two groups. The proportion of patients vomiting was the same in all groups; however, only one patient in the combination group versus three each in the single-agent groups experienced nausea (p = NS). The majority of side effects were associated with the CNS, including somnolence, dizziness, and confusion. Side effects were somewhat more common in both groups receiving dronabinol, though they were not statistically different from the side effects in the group receiving prochlorperazine as a single agent. Efficacy, as measured by duration of nausea and vomiting and by severity of nausea, was significantly greater in both groups receiving dronabinol.