1 Vomiting and restlessness following ENT and eye surgery are undesirable, and may be related to the emetic and analgesic effects of any analgesic given to augment anaesthesia during surgery. 2 To rationalise the choice of analgesic for routine ENT surgery we examined the intraoperative, recovery and postoperative effects following the administration of either buprenorphine (3.0 to 4.5 pg kg-'), diclofenac (1 mg kg-'), fentanyl (1.5 to 2.0 pg kg-'), morphine (0.1 to 0.15 mg kg-'), nalbuphine (0.1 to 0.15 mg kg-'), pethidine (1.0 to 1.5 mg kg-') or saline (as control) given with the induction of anaesthesia in 374 patients. A standardised anaesthetic technique with controlled ventilation using 0.6-0.8% isoflurane in nitrous oxide and oxygen was employed. The study population constituted 7 similar groups of patients. 3 Intraoperatively, their effects on heart rate and blood pressure, airway pressure and intraocular pressure, were similar. This implies, most surprisingly, that neither their analgesic nor their histamine releasing effects were clinically evident during surgery. By prolonging the time to extubation at the end of anaesthesia, only buprenorphine, fentanyl, morphine and pethidine provided evidence of intraoperative respiratory depression. 4 Postoperatively, buprenorphine was associated with severe respiratory depression, prolonged somnolence, profound analgesia and the highest emesis rate. Diclofenac exhibited no sedative, analgesic, analgesic sparing, emetic or antipyretic effects. Fentanyl provided no sedative or analgesic effects, but was mildly emetic. Morphine provided poor sedation and analgesia, delayed the requirement for re-medication and was highly emetic. Nalbuphine and pethidine produced sedation with analgesia during recovery, a prolonged time to re-medication and a mild emetic effect. None provided evidence, from analysis of postoperative re-medication times and analgesic consumption, of any pre-emptive analgesic effect.5 We conclude that nalbuphine (mean dose 0.13 mg kg-') and pethidine (mean dose 1.35 mg kg-'), given individually as a single i.v. bolus during induction of anaesthesia, are the most efficacious analgesics for routine in-patient ENT surgery.