We have compared pain scores at rest and on standardized movement, and morphine consumption using patient-controlled analgesia in 60 patients who had undergone total abdominal hysterectomy. Patients were allocated randomly to one of three groups: in the saline group, 0.9% sodium chloride 50 ml was administered into the pelvic cavity before closure of the peritoneum; in the second group, the solution administered was 20 ml of 0.5% bupivacaine solution with epinephrine 1:200 000 diluted with saline to a final volume of 50 ml; in the third group, the solution used was 20 mg of 2% lidocaine with epinephrine 1:200 000 diluted with saline to a final volume of 50 ml. We found that there was no significant difference between the th ree groups in visual analogue pain scores at 8, 12, 36 or 48 h after operation at rest or on movement, and no significant difference in sedation or dose of antiemetic administered. Mean morphine consumption in the first 24 h was 54.6 (SEM 5.9) mg in the saline group, 55.5 (6.4) mg in the bupivacaine group and 52.5 (5.3) mg in the lidocaine group. In the second 24 h, morphine consumption was 34.9 (6.6) mg, 28.1 (3.5) mg and 28.0 (3.5) mg in the three groups, respectively. We conclude that i.p. administration of local anaesthetic solution into the pelvic cavity did not confer appreciable analgesia in patients undergoing abdominal hysterectomy.