We examined the effects of the following variables on interpleural analgesia after thoracotomy: addition of epinephrine to local anesthetic, thoracostomy drainage, two-catheter placement, and location of catheter tips. Twenty patients were randomized to have one catheter (paravertebral tip location) or two catheters (paravertebral and lateral thoracic wall tip locations). Interpleural catheters were sutured to the parietal pleura by the surgeon at time of wound closure. Patients were then randomly assigned to receive 20 mL of 0.5% bupivacaine with 1:200,000 epinephrine through the single catheter or 10 mL of 0.5% bupivacaine with or without 1:200,000 epinephrine through each of the two catheters while supine. Bupivacaine concentrations in whole blood and in thoracostomy drainage fluid were assayed by gas chromatography. Actual content of bupivacaine in the drainage fluid was calculated. Degree of analgesia was assessed by verbal numerical pain scores over the first 4 h and opioid demand thereafter. Addition of epinephrine to bupivacaine did not influence the degree of analgesia. Approximately 30%-40% of any administered dose of bupivacaine was lost via the thoracostomy tube over a 4-h period. There was no correlation between the true initial dose (100 mg minus thoracostomy drainage) and Cmax. Use of two catheters resulted in significantly less opioid requirements after an initial 8-h period. Failure to achieve adequate interpleural analgesia in postthoracotomy patients may be related to loss of anesthetic via thoracostomy drainage, presence of extravasated blood and tissue fluid in the pleural space, and possibly sequestration and channeling of flow of local anesthetic by restricted motion of an operated lung.(ABSTRACT TRUNCATED AT 250 WORDS)