Sixty-four patients receiving cancer chemotherapy known to induce severe emesis entered a randomized double-blind study of the antiemetic efficacy of haloperidol (Haldol) and benzquinamide (Emetecon). Patients preferred haloperidol for control of emesis induced by cis-platinum (78 vs. 22%) or nitrogen mustard (67 vs. 16%). Patients receiving Doxorubicin preferred benzquinamide by a small margin (46 to 38%). Individual patients who experienced no relief with their first antiemetic (13 of 15) usually got some relief with the other after crossover. Haloperidol was more effective than benzquinamide (54 vs. 29%) in patients previously unrelieved by prochlorperazine (Compazine). Complete relief of vomiting was obtained in 14 of 45 patients receiving haloperidol but only five of 41 patients receiving benzquinamide experienced no vomiting, again dependent on the anticancer agent used. Although haloperidol is a more effective antiemetic agent overall, efficacy is related to the anticancer treatment and probably to individual patient characteristics.