This study evaluated the antiemetic efficacy, cost-effectiveness and clinical utility of prophylactic ondansetron and dexamethasone compared with placebo in the prevention of postoperative nausea and vomiting (PONV) in 135 children (2-15 yr, ASA I-II) undergoing strabismus repair. After induction with halothane and nitrous oxide in oxygen or i.v. thiopental, the children received i.v. dexamethasone 1 mg kg(-1) to a maximum of 25 mg, ondansetron 100 microg kg(-1) to a maximum of 4 mg or placebo (n=45). Episodes of PONV were recorded for the first 24 h after the operation. True outcome measures (parental satisfaction score, duration of stay in the postanaesthesia care unit and fast tracking time), therapeutic outcome measures (number needed to prevent (NNTP) PONV) and the cost to benefit a child with each drug were analysed. The incidence and severity of PONV in the first 24 h were significantly less in the dexamethasone and ondansetron groups than in the placebo group (P<0.05). The incidence (P=0.04) and severity (P=0.03) of PONV at the 6-24 h epoch were significantly less in the dexamethasone group than in the ondansetron group. Recovery time (P=0.07), fast tracking time (P=0.6), parental satisfaction scores (P=0.08) and NNTP PONV were comparable (NNTP=2) in both the ondansetron and the dexamethasone group. The cost to benefit a child with dexamethasone was approximately 22 times less than that of ondansetron.