We tested whether prophylactic droperidol and ondansetron, in combination with a moderate dose of dexamethasone, were equally effective in reducing nausea and vomiting after tonsillectomy in children and that both were superior to saline with dexamethasone. We randomly allocated 300 children to intravenous saline, droperidol 10 μg.kg ^-1 or ondansetron 150 μg.kg ^-1 , after induction of anaesthesia and the administration of intravenous dexamethasone 250 μg.kg ^-1 . The rates (95%CI) of nausea or vomiting within 24 postoperative hours were: 42/91 after saline, 46% (36%-57%); 43/87 after droperidol, 49% (39%-60%); reduced to 18/84 by ondansetron, 21% (13%-32%), p < 0.001. There were no differences in the rates of side-effects between groups. We conclude that ondansetron is more effective than saline in preventing nausea or vomiting after paediatric tonsillectomy when given with a moderate dose of dexamethasone, whereas droperidol was not.