Postoperative pain relief immediately after major surgery cannot be achieved with opioids alone in all patients without respiratory depression or other significant side effects. This investigation was conducted to determine whether the need for opioids and the incidence of side effects can be reduced while maintaining the quality of pain relief using a nonsteroidal antiinflammatory drug as an adjuvant to an opioid. The analgesic efficacy and safety of patient-controlled analgesia using fentanyl with and without intravenous diclofenac were compared after total hip replacement. Forty patients were randomly assigned to receive either diclofenac 75 mg as an initial intravenous loading dose followed by an infusion of 5 mg per hour or saline in a double-blind fashion. The amount of fentanyl administered was recorded. The patients assessed their pain intensity verbally and on a visual analogue scale at intervals of 4 h. The diclofenac group showed a significant reduction in the amount of fentanyl administered during the first 16 h postoperatively as compared to the placebo group (0.65 mg +/- 0.2 vs. 1.08 mg +/- 0.4 respectively, P less than 0.01), and also reported less pain at 16 h (median score on visual analogue scale 0.75 vs. 2.4 respectively, P less than 0.05)). There were no differences in side effects, postoperative blood loss, plasma activated partial thromboplastin time, or Ivy bleeding time between the groups. In conclusion, the addition of diclofenac led to a reduction in fentanyl requirement but did not have any other significant advantages in the treatment of pain following major orthopedic surgery.