We studied 104 patients undergoing primary anterior cruciate ligament reconstruction who were prospectively randomized into receiving either an intravenous pethidine infusion or a continuous bupivicaine infusion to the lumbar plexus via the femoral sheath, for postoperative analgesia. No difference was found between the two methods in terms of their analgesic effect (P > 0.1) and both were well tolerated by the patients with no serious complications. The incidence of nausea, vomiting and drowsiness was significantly higher in the pethidine infusion group (P < 0.05) and remained higher despite the addition of an anti-emetic to the infusion (P < 0.01). The mean pain scores and need for supplementary analgesia were higher in the patients in whom the mini-arthrotomy surgical technique was used when compared with the arthroscopically assisted technique (P < 0.05).