OBJECTIVE To evaluate the efficacy of thoracic epidural analgesia (TEA) in scoliosis surgery. METHODS Descriptive clinical study. METHODS University hospital. METHODS 15 ASA physical status I, II, and III patients undergoing thoracolumbar scoliosis correction. METHODS TEA was performed at three to 5 cm cephalad to the incision, and 5 to 10 mL of 0.125% - 0.2% levobupivacaine was given initially. Then, 5 to 10 mL of levobupivacaine was infused hourly throughout the operation. General anesthesia (GA) was induced with thiopental sodium (5 mg/kg) and fentanyl (one μg/kg) and was maintained with 0.2% sevoflurane and 50% nitrous oxide in oxygen. Intraoperative epidural morphine (two to three mg) was administered, and 0.1% levobupivacaine with morphine (0.04 to 0.08 mg/mL) was infused at two to 4 mL/hr for postoperative analgesia. METHODS Adequacy of anesthesia, postanesthetic recovery and analgesia, adverse effects, and patient satisfaction were recorded. RESULTS 20% of patients underwent more than 10 levels of correction, and 53% had coexisting morbid diseases. All had adequate anesthesia. Immediately in the Postanesthesia Care Unit (PACU), 67% of patients reached an Aldrete score of 10, and 40% were fully awake and oriented. All patients were arousable to command and able to flex their hips and knees. None had intraoperative recall. 73% reported no pain in the PACU or 6 hours postoperatively. No serious adverse effects occurred. 80% of patients rated their satisfaction as "good". CONCLUSIONS Preincisional application of TEA with light GA may be used effectively in thoracolumbar scoliosis surgery.