We have compared the efficacy of ondansetron, metoclopramide, droperidol and placebo in the prevention of postoperative nausea and vomiting in 118 day stay patients undergoing laparoscopic gynaecological procedures. All received a standardised general anaesthetic of fentanyl, propofol, nitrous oxide in oxygen and isoflurane. Three to five min before induction of anaesthesia, patients were allocated to receive ondansetron 4 mg, metoclopramide 10 mg, droperidol 1 mg or placebo in a randomised, double-blind manner. Visual analogue scores for nausea, the incidence of emetic episodes, and analgesic and antiemetic consumption were recorded for 48 h postoperatively. The scores for nausea were significantly lower in the ondansetron group (p < 0.01) compared with the other three groups at 1, 2 and 4 h after operation; thereafter there was no difference. The incidence of emesis was lower (p = 0.063) and time to first oral fluids was shorter (p < 0.05) in the ondansetron group. Oral analgesic requirements were significantly greater in the ondansetron group over the 48 h study period. Two patients, one each in the placebo and metoclopramide groups, had to remain in hospital overnight because of persistent emetic symptoms.