Background: Isoflurane has exceeded halothane and enflurane in usage. A literature search, however, revealed no data comparing the effects on emesis, headache and restlessness of these three agents. Methods: With hospital ethics committee approval and patient consent, a prospective, randomised, double-blind study of 556 patients undergoing ENT and eve surgery was undertaken to evaluate the effects of halothane, isoflurane and enflurane on vomiting, retching, headache and restlessness until 24 h after anaesthesia. Balanced general anaesthesia was administered comprising benzodiazepine premedication, induction with thiopentone-atracurium-morphine (ENT patients) or fentanyl (eye patients), controlled ventilation and maintenance with either halothane 0.4-0.6 vol% (n=186), isoflurane 0.6-0.8 vol% (n=184) or enflurane 0.8-1 vol% (n=186) in nitrous oxide 67% and oxygen. Results: ?The three study groups were comparable, and comprised comparable subgroups having ear, nose, throat, intraocular and non-intraocular surgery. During early recovery from anaesthesia, the respective requirements for halothane, isoflurane and enflurane for analgesia (7%, 9% and 10%), frequency of emesis (6%, 8% and 8%), antiemetic requirements (1%, 1% and 2%), restlessness-pain scores and time spent in the recovery ward (27 SD 10, 31 SD 12 and 26 SD 9 min) were similar. During the ensuing 24-h postoperative period, patients who had isoflurane experienced emesis less often than those who had halothane (36% vs 46%, P<0.025) but did so with similar frequency to those who had enflurane (46% vs 41%). Antiemetic requirements were least in those given isoflurane (isoflurane 12%, halothane and enflurane 23% each, P<0.005), but headache and analgesic requirements were similar. Conclusion: Isoflurane induces less postoperative emesis than halothane, but headache is similarly frequent after anaesthesia with any of these agents.