We studied 100 ASA I-II females undergoing general anaesthesia for major gynaecological surgery, in a prospective, double-blind, placebo-controlled, randomized study. Patients received one of four regimens for the prevention of postoperative nausea and vomiting (PONV): ondansetron 4 mg (n = 25), dexamethasone 8 mg (n = 25), ondansetron with dexamethasone (4 mg and 8 mg, respectively, n = 25) or placebo (saline, n = 25) There were no differences in background factors or factors related to operation and anaesthesia, morphine consumption, pain or side effects between groups. The incidence of nausea and emetic episodes in the ondansetron with dexamethasone group was lower than in the placebo (P < 0.01), ondansetron (P < 0.05) and dexamethasone (P = 0.057) groups. There were no differences between ondansetron and dexamethasone, and both were more effective than placebo (P < 0.05 and P < 0.01, respectively). Dexamethasone appeared to be preferable in preventing nausea than emetic episodes. Fewer patients in the ondansetron with dexamethasone group needed antimetic rescue (P < 0.01 vs placebo and P < 0.05 vs ondansetron). We conclude that prophylactic administration of combined ondansetron and dexamethasone is effective in preventing PONV.