In order to further examine the possible role of endogenous opioid peptides and vasopressin in the phenomenon of dexamethasone nonsuppression, we studied the effect of naloxone, vasopressin, and vasopressinnaloxone combination on cortisol secretion following dexamethasone pretreatment. Nine healthy males were given 1 mg dexamethasone at 23.00 h. The following day starting at 12.30 h and at 90-min intervals they received intravenously naloxone (0.2 mg/kg), arginine vasopressin 3 units, or the two drugs combined. The order of drug administration was counterbalanced using a Latin square design. Blood samples were drawn at 15-min intervals, and plasma aliquots were assayed for cortisol and dexamethasone. Naloxone failed to induce an escape from dexamethasone suppression. Four of the 9 subjects responded with an escape from dexamethasone suppression in response to vasopressin alone. The observed variability in response to vasopressin was unrelated to dexamethasone plasma levels but was associated with a decrease in systolic blood pressure. Peak cortisol levels were lowest in response to naloxone and highest in response to vasopressin. There was no evidence of an increased cortisol response to the coadministration of naloxone with vasopressin compared to vasopressin alone. These results fail to implicate an opioidergic mechanism in the pathophysiology of dexamethasone nonsuppression.