The benefits of epidural narcotic analgesia (ENA) have been documented in mixed surgical populations. To assess the safety and utility of ENA after thoracic surgery and to assess potential interactions with intraoperative intravenous narcotics (IIN), we retrospectively examined the records of 130 consecutive patients having thoracotomy. The 116 patients who received ENA required a mean of 0.19 mg/kg of intravenous morphine sulfate (MS) within the first 48 postoperative hours, as opposed to 0.44 mg/kg for patients who did not receive ENA. The place in which nonepidural patients were extubated most frequently was the operating room (71%), followed by the intensive care unit (21%) and the recovery room (7%). Percentages were similar for epidural patients: 71% were extubated in the operating room, 20% in the intensive care unit, and 9% in the recovery room. Nonepidural patients had an immediate mean postoperative PCO2 of 39.2 mm Hg, epidural patients a mean of 40.1 mm Hg. There were no technical complications due to epidural catheter placement, and no reintubation was required within the first 72 postoperative hours. The concomitant administration of IIN did not produce a significant difference in postextubation PCO2 in either group of patients, although increasing doses resulted in a lower percentage of patients extubated in the operating room or recovery room. We conclude that ENA may be safely administered to patients having thoracotomy, and it diminishes the need for postoperative intravenous narcotics.