The 5-HT3 antagonists are effective in reducing postoperative nausea and vomiting (PONV) associated with paediatric tonsillectomy. Although prophylactic tropisetron can reduce the incidence of PONV by half, the resulting level of over 40% is still unacceptably high. The aim of this study was to evaluate the effect of adding dexamethasone to tropisetron. In a blinded study, 59 children (mean age 6.1 years) were administered 0.1 mg.kg-1 up to 2 mg of tropisetron and 66 children (mean age 5.7 years) received the same dose of tropisetron plus 0.5 mg.kg-1 up to 8 mg of dexamethasone. Both drugs were given intravenously during induction of anaesthesia for tonsillectomy. During the inpatient stay of 24 h, the incidence of postoperative vomiting in the tropisetron alone group was 53% compared with 26% in the combination group (P=0.002, chi-squared). A significant reduction in nausea from 53% to 30% was also observed (P=0.02). Parents completed a daily diary for 5 days following discharge. Delayed vomiting occurred in 27% and 11% of the tropisetron and combination therapy groups, respectively (P=0.025) Sixteen percent and 9%, respectively, required medical attention (P=0.27). Tropisetron plus dexamethasone is more effective than tropisetron alone in reducing the incidence of PONV following paediatric tonsillectomy.