Improved control of postoperative pain is now known to reduce the incidence of morbidity. Although spinally administered narcotics have found a clear role in chest and abdominal surgery, their role in lumbar spinal surgery is debated. We conducted a prospective, double-blind, randomized, placebo-controlled trial of intrathecally administered morphine sulfate after lumbar spinal surgery in 56 patients. Patients received 0, 0.125, 0.25, or 0.5 mg of intrathecally administered morphine during extradural lumbar spinal operations, and the effects on postoperative analog pain scores, narcotic consumption, complications, and length of hospitalization were assessed. As compared with systemic narcotic administration, intrathecally administered morphine provided superior analgesia in a dose-dependent fashion without an increase in narcotic side effects. Consumption of parenteral narcotics on the first postoperative day and over the total hospitalization period decreased in correlation with increasing doses of intrathecally administered morphine. Mean length of hospitalization was significantly decreased, as compared with the control group, in patients receiving 0.25 or 0.5 mg of intrathecally administered morphine. When proper precautions are observed, intrathecally administered morphine can improve the postoperative care of patients undergoing lumbar spinal surgery.