In addition to classic somatic signs of opiate withdrawal, a number of behavioral measures are known to be sensitive, reliable indices of naloxone-precipitated opiate withdrawal in rats. It has been suggested that some behavioral indices of withdrawal may be more sensitive to precipitation by naloxone than some somatic signs of withdrawal. The purpose of the present study was to permit a quantitative assessment of the relative sensitivity to naloxone of a variety of behavioral and somatic indices of opiate withdrawal. Male Wistar rats were implanted s.c. with either two morphine (each 75 mg of base) or two placebo pellets. No sooner than 3 days after implantation, naloxone dose-response functions were determined with several behavioral paradigms and ratings of a variety of somatic withdrawal signs. In dependent rats, very low (0.004 or 0.01 mg/kg) doses of naloxone produced the following behavioral effects: 1) a reduction in spontaneous locomotor activity, 2) a disruption of schedule-controlled (fixed ratio 15) operant responding for food, 3) an elevation in intracranial self-stimulation thresholds and 4) a conditioned place aversion. These same doses of naloxone produced no significant effects in nondependent (placebo pellet-implanted) rats. The ED50 values for naloxone precipitation of all behavioral signs of withdrawal were below 0.013 mg/kg; the ED50 values for naloxone precipitation of most somatic withdrawal signs were higher. The behavioral measures used in these studies therefore represent highly sensitive indices of opiate withdrawal.