Background: The highly lipid soluble opioid, fentanyl, has a rapid onset and short duration of action. The present study was designed to examine the analgesic efficacy and side effects of the combination of fentanyl with morphine in patients using intravenous PCA. Methods: Patients were randomly assigned to receive one of three PCA regimens: M4 group (40 mg morphine+90 mg ketorolac+1.5 mg dorperidol), M2F2 group (20 mg morphine+200 ug fentanyl+90 mg ketorolac+1.5 mg dorperidol), or M2F4 group (20 mg morphine+400 ug fentanyl+90 mg ketorolac+1.5 mg dorperidol). All patients were given initial loading dose of 0.1 mg/kg morphine plus 1 mg droperidol at the end of surgery. Pain score, side effects, and overall satisfaction were assessed at 30 min, 1 hr, 8 hr, 24 hr, and 48 hr postoperatively. Results: The pain score was significantly higher in the M2F2 group than in the M4 group and M2F4 group during 1 hr and 8 hr postoperatively. The total opioid consumption was significantly greater in the M2F4 group than in the M4 group. Patient satisfaction was better in the M2F4 than other two groups. There were no differences in the overall incidence of side effects among three groups. Conclusions: The present results suggest that the combination of fentanyl with morphine for intravenous patient-controlled analgesia is a useful method, and the double dose of fentanyl in comparison with the equipotent morphine dose is recommended in the early postoperative period.