BACKGROUND: Thoracic epidural block has been widely used to control postoperative pain. The relative effects of the mass, volume, and concentration of local anesthetic solutions used are still subjects of debate, and the relationship between general anesthesia and thoracic epidural block is unclear. METHODS: In study I, 20 conscious patients scheduled for elective upper abdominal surgery received 0.2% bupivacaine with morphine 3 mg (group CL: n = 10) or 0.5% bupivacaine with morphine 3 mg (group CH: n = 10) via an epidural catheter inserted at the T8-T9 epidural space. The width of sensory block, blood pressure and heart rate were assessed for 30 minutes and then general anesthesia was administered. In study II, under general anesthesia, 72 patients scheduled for elective upper abdominal surgery were injected with the same formula as used in study I (group UL: n = 36, group UH: n = 36). Blood pressure, heart rate, vol% of end-tidal inhalation agent were measured for an hour. Postoperatevely, the VAS scores of both study groups were assessed in the recovery room, after 24 hours and 48 hours. RESULTS: In study I, the mean arterial pressures were significantly lower in the CH group than the CL group (P = 0.0419 and P = 0.0328 at 20 and 25 minutes after local anesthetic infusion), but heart rate and VAS score were not significantly different. In study II, no difference was observed between the groups in terms of blood pressure, heart rate, enflurane requirement, or VAS score. But hypotension requiring treatment was more frequent in the UH group than in the UL group (group UL: n = 3, group UH: n = 6). CONCLUSIONS: Epidural administration of the same volume of 0.2% bupivacaine or 0.5% bupivacaine provided similar pain relief and hemodynamic change, but profound hypotension was more frequent in the 0.5% bupivacaine group during general anesthesia. This study shows no benefit of using 0.5% bupivacaine to improve postoperative analgesia in upper abdominal surgery.