BACKGROUND Morphine has been the standard opioid in patient-controlled analgesia (PCA). Oxycodone, the analgesic potency of which in i.v. administration has been suggested to be slightly greater than that of morphine, has not yet been studied for its efficacy in PCA. METHODS Fifty patients, undergoing a plastic reconstruction of the breast or a major operation of the vertebrae, such as lumbar spinal fusion, used PCA for postoperative pain. Patients were randomized to receive either morphine 45 microg/kg or oxycodone 30 microg/kg as i.v. bolus doses. Patients were assessed for pain with a visual analogue scale (VAS) and side effects at 3, 9 and 24 h. Venous blood samples for the measurement of plasma concentration of oxycodone and that of morphine and its metabolites were taken. RESULTS In this study patients needed, on average, the same amount of oxycodone and morphine in the recovery room and on the ward. There was no difference in the quality of analgesia (VAS) or incidence of side effects, such as nausea, vomiting, pruritus and urinary retention. The plasma concentrations of morphine-6-glucuronide showed that this metabolite might contribute to the analgesia resulting from morphine administration. CONCLUSIONS The same dose of intravenous oxycodone and morphine administered by PCA pump was needed for immediate postoperative analgesia. The two drugs appear to be equipotent.