Background/Aim. Combined spinal-epidural-general anesthesia has several advantages over general anesthesia alone. This study was designed to compare the efficacy of intrathecal (IT) morphine alone, or in combination with bupivacaine and fentanyl, as part of a combined spinal-epidural (CSE) analgesia, in patients undergoing elective total gastrectomy. Methods. This prospective, randomized double-blind study included 60 patients undergoing total gastrectomy under general anesthesia and CSE. We compared the analgesic effect of lumbar IT morphine 300 μg (the group M, n = 20) vs morphine 300 μg + bupivacaine 2 mg (the group MB, n = 20) vs morphine 300 μg + bupivacaine 2 mg + fentanyl 25 μg (the group MBF, n = 20) given after thoracic epidural catheter placement (T6-7) but before general anesthesia induction. Pain visual analogue scale (VAS) at rest (R), with movement (M) and with cough (C), and the number of analgesia requests were assessed for 72 h and after epidural catheter removal. Results. Compared to other groups, the MBF group required significantly fewer additional intra-operative epidural bupivacaine doses (p < 0.001), whereas the M group required significantly more supplemental intraoperative intravenous fentanyl, compared with the MBF (p = 0.022) and MB groups (p = 0.005). Postoperative pain relief was satisfactory in all the groups at all the time. VAS-R and VAS-M did not differ significantly among the groups. Compared to the M group, VAS-C scores 30 min postoperatively were significantly lower in the MBF (p = 0.029) and MB groups (p = 0.002). Duration of analgesia was longer in the MBF and MB groups, but the difference reached no significance. The number of supplemental analgesia requests was similar in all the groups in the first 12 h and during 72 h. Additional analgesia requests after epidural catheter removal were similar in all the groups, and side effects were infrequent. Conclusion. Compared to IT morphine alone, triple IT combination administered as part of CSE provided better intraoperative analgesia, but conferred no benefit with regards to postoperative analgesia.