This randomized, double-blind study compared the efficacy of two mu-receptor antagonists, naloxone and nalbuphine, in the prophylactic management of pruritus in postcesarean section patients receiving epidural morphine. Dosages of study drugs were individualized by the use of a patient self-administration (PSA) device. All 51 patients were healthy women who received a uniform epidural anesthetic and epidural morphine (5 mg). Coded solutions were infused for 24 h, with 5-min PSA lockout times: Group A (n = 17), nalbuphine 2.5 mg/h, PSA nalbuphine 1 mg; Group B (n = 16), naloxone 50 micrograms/hr, PSA saline; Group C (n = 18), naloxone 50 micrograms/h, PSA naloxone 40 micrograms. Patients were assessed for pruritus and pain every 8 h for 24 h. Both naloxone and nalbuphine provided good relief for pruritus; median pain and pruritus scores were in the none-to-mild range (0-3) for all groups at all assessment intervals. The pruritus scores of the PSA saline group were higher during the 16- to 24-h period (P < 0.05) than the scores of either group receiving A-receptor antagonist by PSA. There was evidence of shortening of the duration of analgesia in patients receiving naloxone who required treatment for pruritus after 16 h. Patients who self-administered large doses of nalbuphine over the first 8 h also reported pain scores consistent with reversal of analgesia. The potency ratio for naloxone:nalbuphine for antagonism of the pruritic effects of epidural morphine was approximately 40:1. Intervention to treat either unrelieved pruritus or pain, respectively, was necessary in the following numbers of patients: Group A, 0/1; Group B, 1/1; Group C, 2/2. Prophylactic infusions offer the potential for labor cost savings by minimizing the need for episodic therapeutic interventions to treat pruritus.