The influence of two different concentrations of iv naloxone infusion on the analgesia and adverse effects of epidural morphine were compared in a double-blind, placebo-controlled study. Fortyfive patients undergoing gallbladder surgery were provided postoperative analgesia by 4 mg epidural morphine; they then received an iv infusion over a 12-h period consisting of either 5 μg · kg−1· h−1naloxone, 10 μg · kg−1· h−1naloxone, or saline. Pain relief was assessed by hourly visual analog scoring (VAS) and by direct questioning of the patient. Requirement of additional analgesia was noted. Respiratory frequency was monitored every 15 min and arterial blood gases were analyzed every 2 h for 24 h. Peak expiratory flow (PEF) was recorded 6 and 24 h postoperatively. Steady-state kinetics of naloxone were determined by a modified radioimmunoassay (RIA) method. All patients had good to excellent postoperative pain relief. Naloxone, 5 μg · kg−1· h−1, did not appear to have any effect on epidural morphine analgesia. However, naloxone infusion at the rate of 10 μg.kg−1h−1reduced the duration of analgesia by about 25%, and more frequent injections of epidural morphine were required to give effective analgesia. Complete reversal of analgesia was not seen in any patient. A dose-related stimulatory effect on respiratory frequency was noted in the groups receiving naloxone. Paco2values also were better in these groups as compared to values in the placebo group. The steady-state plasma concentration of naloxone was 2.8–3.7 ng/ml during infusion at the rate of 5μg.kg−1h−1and 4.3–5.1 ng/ml during 10 μg · kg−1h−1naloxone infusion. The plasma clearance of naloxone was 30.5 and 35.4 ml. min−1. kg−1for the low and high dose groups, respectively, and showed a four-fold interindividual variation. The authors conclude that naloxone reverses epidural morphine analgesia in a dose-dependent manner. Low-dose naloxone infusion (5 μg.kg−1h−1)prevents respiratory depression due to epidural morphine without affecting its analgesia.