Forty dogs undergoing a variety of surgical procedures were assigned randomly to one of two groups. All the animals were premedicated with acepromazine (0.05 mg/kg bodyweight) intramuscularly, and anaesthesia was induced with thiopentone sodium, or propofol in the case of lean animals, and maintained with halothane in an oxygen/nitrous oxide mixture using a non-rebreathing circuit. The dogs in group 1 were given papaveretum (0.2 mg/kg) slowly intravenously within 35 minutes of induction of anaesthesia and the dogs in group 2 were given carprofen (4 mg/kg) in the same way. The dogs were scored for sedation and pain by a trained theatre nurse, who did not know which group they belonged to, using a visual analogue scale, at 15, 30, 60, 120, 240 and 360 minutes after the halothane was switched off at the end of the procedure. Nine of the dogs were withdrawn from the trial (eight of them from the papaveretum group) because of inadequate pain relief and these animals were given pethidine (3 mg/kg intramuscularly) which produced adequate analgesia within 15 minutes in all but one case. Carprofen provided profound analgesia which was as effective and of longer duration than that produced by papaveretum, and was associated with significantly less postoperative sedation and a quicker return to the normal conscious state.