The safety and efficacy of ondansetron were evaluated in the treatment of postoperative nausea and vomiting. Five hundred patients who experienced nausea or vomiting in the Post-Anaesthesia Care Unit within the first 2 h of recovery were randomized to receive either 1, 4, or 8 mg of ondansetron, or placebo. All patients had undergone ambulatory surgery with general endotracheal anaesthesia. Episodes of emesis, nausea scores, adverse events, vital signs, and laboratory values were assessed before and during the 24 h after study drug administration. Patients were evaluated for the first 2 h in the Post-Anaesthesia Care Unit then followed up for the next 22 h. Complete response was defined as no emetic episodes, no nausea or no rescue anti-emetic medication. For the 0-24 h study period, complete response occurred in only 15% of the placebo group compared to 41%, 47%, and 47% in the 1, 4, and 8 mg ondansetron groups, respectively. Mean nausea scores (scale of 0-10) during the initial observation period (0-2 h) were significantly lower for all doses of ondansetron [2.2 (1 mg), 1.7 (4 mg), and 2.1 (8 mg)] compared to placebo (3.0). The optimal dose of ondansetron for the treatment of postoperative nausea and vomiting was found to be 4 mg. All doses of ondansetron were well tolerated. No clinically significant increases in laboratory parameters or alterations in haemodynamic stability occurred in the ondansetron groups compared to placebo.