There have been conflicting reports of the value of naloxone infusions to prevent the side-effects associated with epidural morphine. In our study, 29 patients undergoing thoracotomies for pulmonary surgery received epidural morphine (0.1 mg.kg-1) shortly after induction of anaesthesia. One hour after arrival in the Recovery Room, one of four naloxone bolus and infusion sequences was administered: saline bolus followed by saline infusion; 0.4 microgram.kg-1 naloxone bolus followed by 0.4 microgram.kg-1.hr-1 naloxone infusion; 2.0 micrograms.kg-1 naloxone bolus followed by 2.0 micrograms.kg-1.hr-1 naloxone infusion; and 4.0 micrograms.kg-1 naloxone bolus followed by 4.0 micrograms.kg-1.hr-1 naloxone infusion. Although with the number of patients studied, there were no statistically significant differences among groups, clinically, there was a trend toward decreased analgesia with all three naloxone infusion doses as determined by analgesic requirements, longest analgesic-free period and visual analogue pain scores. In addition, side-effects occurred in all groups. We conclude that prophylactic naloxone, used in this manner, is not an appropriate technique for the prevention of side-effects associated with epidural morphine used for the prevention of pain after thoracotomy.