Background: Preoperative blocking of surgical nociceptive inputs may prevent sensitization of CNS and reduce postoperative pain. The stress responses to surgical trauma consist of increase in catabolic hormones and decrease in anabolic hormones. We studied whether preoperative intravenous morphine could affect postoperative pain and change plasma cortisol and serum glucose levels. Methods: Thirty eight patients undergoing total abdominal hysterectomy were randomly assigned to one of three groups. Control group (n=11) did not received intravenous morphine, preoperative group (n=13) received intravenous morphine (0.1 mg/kg as a bolus 10 min before operation and followed by 1.5 mg/hr for 10 hours), postoperative group (n=14) received the same doses and method of intravenous morphine of preoperative group postoperatively. Postoperative pain relief was provided with i.v. fentanyl through Patient-Controlled-Analgesia Pump. Postoperative visual analogue scores (VAS), analgesic requirement (first request time, total amounts used), side effects, plasma cortisol and serum glucose levels were compared. Results: VAS were different between control group and the other two goups, but were not different between preoperative and postoperative group. Total amounts of used fentanyl were not different among groups, but first request time were significantly delayed in the preoperative group compared with the other two groups ($66.2{pm}33.9$ vs $39.0{pm}15.4$ and $45.0{pm}14.9$ min respectively, p<0.05). Plasma cortisol and serum glucose levels were not different among groups. Conclusions: Above dosage of preoperative and postoperative morphine has analgesic effect, but could not block surgical stress induced plasma cortisol and serum glucose increase.