In order to assess the analgesic properties of epidural low-dose morphine and its possible influence on the adrenocortical, hyperglycemic, renal, electrolyte and leukocyte responses to surgery and nitrogen excretion, a double-blind randomized study was undertaken in 14 otherwise healthy patients admitted for hysterectomy under halothane, N2O/o2 anesthesia. Before induction of anesthesia, an epidural catheter was introduced into the lumbar epidural space. After induction of anesthesia, either morphine 4 mg in 10 ml saline or 10 ml saline was injected into the epidural space, according to the allocation. Postoperatively, the degree of pain was evaluated by mean of a visual analogue scale (0-10). When pain score exceeded 5 points during the 24 -h trial, either 4 mg morphine in saline or saline was given epidurally. If the pain score did not decrease more than 2 points after an epidural injection, morphine was given parenterally (5 mg i.v. +5 mg i.m.). The results showed that pain scores, duration of pain relief and doses of morphine differed significantly between groups (P less than 0.05). Plasma concentration of cortisol and glucose, plasma-and urine electrolytes, 24-h creatinine and free-water clearances, diuresis, fluid balance, leukocyte count and nitrogen excretion differed insignificantly between groups. In conclusion, epidural low-dose morphine is a superior alternative to conventional postoperative pain treatment because of greater and longer lasting pain relief, without apparent side-effects. The measured endocrine-metabolic and renal response did not differ between groups, indicating that low-dose epidural morphine does not inhibit afferent neurogenic stimuli from the site of surgical trauma.