The intrapleural administration of local anesthetics via a catheter is a relatively new method for postoperative analgesia after thoracic and upper abdominal surgery. Many authors have described this technique as effective and with few side effects. METHODS. Intrapleural administration of bupivacaine (IPC group) was compared with intramuscular administration of buprenorphine in 26 patients who had undergone a cholecystectomy. During the 1st postoperative week the pulmonary function (vital capacity, forced expiratory volume in 1 s [FEV1]) and an arterial blood gas analysis were performed daily and the quality of analgesia was evaluated. RESULTS. The patients in the two groups were similar with regard to sex, age, weight, and height. The postoperative arterial oxygen partial pressure (pO2) did not differ significantly from the preoperative value in the IPC group, whereas in the opioid group the results were significantly lower than preoperatively (p less than 0.05 6 h after the operation and p less than 0.005 from the 1st to 7th postoperative days). There was a significantly higher pO2 in the opioid group preoperatively that, however, had no correlation with the differences between pre- and postoperative values. After the operation, the differences from the preoperative value were significantly smaller in the opioid group from the 1st to 7th postoperative days, reaching a significance of p less than 0.001 on the 2nd and 3rd, postoperative day. The pCO2 was slightly increased after the operation, but there were no significant differences between the groups. A significant decline in vital capacity and FEV1 occurred in both groups during the first days after the operation; there were no significant differences between the two groups during these days. The degree of pain was similar in the two groups before the analgesic was given. In both groups there was significant improvement of the degree of pain after administration of the analgesic. One hour after the injection, however, the degree of pain was significantly lower in the IPC group compared with the opioid group. Postoperatively, pathologic findings could be seen more often on chest X-rays in the opioid group, although there was no significant difference between the groups. There were no side effects caused by the intrapleural catheter or the local anesthetic except for a minor pneumothorax in 1 patient that did not need any treatment. CONCLUSION. The intrapleural administration of local anesthetic provided very good analgesia in our study. The pO2 was significantly higher and the analgesia was significantly better than after intramuscular opioid administration. Moreover, the technique is simple to perform and has few side effects. We therefore believe it should be employed for postoperative analgesia after thoracic and upper abdominal surgery as well as for other indications to a greater extent.