It has been postulated that chronic pain over-stimulates the hypothalamus-pituitary-adrenal axis to produce an extended stress response. If this assumption is true, patients with severe, chronic pain should demonstrate abnormalities of the pituitary-adrenal axis. To evaluate this premise, we screened 40 adult, chronic pain patients in the first week of treatment with serum cortisol concentrations taken between 8:00 and 10:00 AM. Criteria for inclusion in this study required that pain be present for at least one year and be constant, incurable, interfere with sleep, and cause the patient to be bed or house-bound without opioid treatment. Sixteen (16) of the subjects were challenged with cosyntropin, 0.25mg, given intramuscularly immediately after blood was drawn for determination of baseline serum cortisol concentration. Normal serum cortisol concentrations was considered to range from 5.0 to 25.0 ug/dl at baseline, and normal cortisol reserve was considered to be at least a doubling of the baseline concentration determined one hour after cosyntropin administration. Ten (25%) of the patients had evaluated serum cortisol concentrations above 25ug/dl with the highest being 54.4 ug/dl. Nine (20%) demonstrated low serum cortisol concentrations under 5 ug/dl, and 5 of 16 (31.25%) given cosyntropin challenge demonstrated inadequate adrenal reserve by failing to double their baseline cortisol concentration. All patients with high or low serum cortisol concentration demonstrated a normal cortisol concentration following pain control with a long-acting opioid including methadone, extended-length morphine or oxycodone, or transdermal fentanyl. This study suggests that severe, chronic pain may produce profound abnormalities of serum cortisol and cortisol reserve, and normalization of these alterations may require pain treatment with long-acting opioids.