The effects of neuroleptanaesthesia on endocrine-metabolic changes during elective gastrectomy was investigated, comparing with those of epidural anaesthesia. Nine patients were given neuroleptanaesthesia and fifteen patients given thoracic epidural analgesia combined with general anaesthesia. We evaluated the levels of stress hormones, insulin and blood glucose. Epidural anaesthesia suppressed the increase of catecholamine, but neuroleptanaesthesia did not inhibit the elevation of the catecholamines. In neuroleptanaesthesia group, glucagon and growth hormone increased during surgery, and the levels of these hormones were significantly higher than those of epidural analgesia group. Blood glucose increased during operation in both groups. In epidural anaesthesia, the levels of insulin and insulin/glucose ratio were kept higher than those of neuroleptanaesthesia group, but this was not statistically significant. Lactate/pyruvate ratio and free fatty acid did not show any significant change during the study in both groups. These results suggest that neuroleptanaesthesia is not a suitable method for upper abdominal surgery.