Postoperative pain relief was studied in 280 patients undergoing various kinds of surgery, e.g., thoracic, upper and lower abdominal, perineal, obstetric, and orthopedic. Morphine, 2 or 4 mg, was given after surgery through an indwelling epidural catheter. Excellent analgesia was noted in 87% of patients; only 3.5% of patients were dissatisfied. A single injection gave complete pain relief for the entire postoperative period in 30% of cases; in the remaining patients the mean duration of analgesia was 10.7 hours (SD ± 4.3). Plasma morphine concentrations recorded after 2-mg doses suggest a regional spinal action as the basis for the long duration of analgesia, although the initial effect after 4-mg doses might well include systemic responses due to rapid vascular uptake of morphine from the epidural space. Peak expiratory flow (PEF) measurements and arterial blood gas analyses showed no significant early postoperative respiratory depression. Absence of sedation, orthostatic hypotension, respiratory depression, and motor paralysis facilitated early ambulation with less risk for postoperative respiratory complications. It is concluded that 2-mg doses of epidural morphine give good analgesia of long duration despite low plasma levels. After upper abdominal and thoracic surgery higher doses (4 mg) may be necessary in healthy patients. Elderly and frail patients appear to be sensitive to epidural morphine and doses in excess of 2 mg should be avoided regardless of the type of surgery. With this dose schedule we have not encountered delayed respiratory depression.