The 100 inescapable tail-shock paradigm produces three sequential analgesic states as the number of shocks increases: an early opioid analgesia (after 2 shocks) that is attenuated by systemic naltrexone, a middle analgesia (after 5-40 shocks) that is unaffected by systemic naltrexone, and a late opioid analgesia (after 80-100 shocks) that is attenuated by systemic naltrexone. In order to determine whether the absence of adrenal hormones would affect any of these analgesias, we tested adrenalectomized (ADX) versus sham-operated control rats 2 weeks post-surgery. Pain threshold was assessed using the tail-flick (TF) test. ADX attenuated both the early (2 shock) and late (80-100 shock) opiate analgesias and failed to reduce the naltrexone-insensitive analgesia after 5-40 shocks. We demonstrated that a loss of adrenomedullary catecholamines does not underlie the ADX-induced attenuation of opioid analgesia since sympathetic blockade using systemic chlorisondamine (6 mg/kg) failed to reduce analgesia at any point in the shock session. It was further shown that stress levels of adrenal hormones are not critical since (a) analgesia was unaffected when animals were tested 48 h after ADX, (b) 2 shocks do not produce a surge in corticosterone (CORT) over and above levels observed in animals restrained and TF tested in preparation for shock, and (c) basal CORT replacement in drinking water fully restored analgesia in ADX rats. These experiments demonstrate that basal CORT, rather than adrenomedullary substances, is critical to the expression of analgesia. The function of CORT here is not linked to a shock-induced surge of the steroid. CORT appears to play a permissive role in the expression of analgesia. Potential effects of the absence of corticosteroids on neurotransmitter biosynthesis important in analgesia production are discussed.