Three doses of intravenous (i.v.) ondansetron, 1 mg, 4 mg, and 8 mg, were compared to placebo for their antiemetic effect and safety. The drugs or placebo were administered in a double-blind manner, prophylactically to 589 women undergoing elective outpatient surgical procedures under nitrous oxide opioid-based general endotracheal anesthesia. In the postanesthesia care unit, the number of emetic episodes, periodic assessments of nausea severity using an 11-point scale (0 = no nausea; 10 = worst nausea), vital signs, and adverse events were collected by an independent observer for 2 h. Upon discharge, identical information, with the exception of vital signs, was collected from the patients' diary and via phone call. One pre- and two poststudy blood specimens for hematology and chemistries were evaluated. During the initial 2 h, patients receiving any dose of ondansetron had significantly better complete response rates (no emesis) than those receiving placebo. Over the 24-h study period, patients who received either 4 mg or 8 mg ondansetron continued to have significantly greater complete response rates. Adverse events were minor, and ondansetron-treated patients had profiles similar to those of the placebo. Heart rate, blood pressure, respiratory rate, and laboratory safety variables were not different among the groups. Ondansetron did not prolong awakening time. This study indicates that ondansetron is a safe and effective prophylactic antiemetic for women who have outpatient surgery under nitrous oxide opioid-based general anesthesia.