This study was designed to compare the efficacy of subarachnoid morphine alone or in combination with bupivacaine and fentanyl for combined spinal-epidural analgesia in colorectal surgery. This is a prospective, randomised, double-blind clinical trial. Sixty patients undergoing low anterior resection were assigned to one of three groups: subarachnoid morphine, bupivacaine and fentanyl, subarachnoid morphine and bupivacaine or subarachnoid morphine only. Epidural catheter placement and subarachnoid injection were done via a combined spinal-epidural Epistar needle at L2-3. The epidural catheter was used for scheduled intraoperative bupivacaine and intermittent postoperative bupivacaine and morphine administration. Intraoperative epidural bupivacaine, intraoperative intravenous fentanyl use, time to first analgesia request, postoperative visual analogue scale pain scores, tramadol requirements and side-effects were recorded for 72 hours. Postoperative analgesia was comparable in all groups. Intraoperative fentanyl and bupivacaine consumption was lowest in the morphine, bupivacaine and fentanyl group. Time to first analgesia request was longer in the morphine, bupivacaine and fentanyl compared to the morphine group (P = 0.009). Tramadol use was lower in the morphine and bupivacaine group compared to morphine, bupivacaine and fentanyl (P = 0.017) on postoperative day two. There were no significant adverse effects. All patients ambulated the morning after surgery. The addition of bupivacaine and fentanyl to subarachnoid morphine did not confer any advantage on postoperative visual analogue scale scores and tramadol use, but lowered the need for additional intraoperative intravenous fentanyl and epidural bupivacaine and prolonged the time to first analgesia request.