Experimental evidence indicates that arginine vasopressin contributes to the release of adrenocorticotropic hormone under certain conditions. We studied for the first time the AVP antagonist [d(CH2)5 Tyr(Me)AVP] in 6 normal men in order to evaluate the possible role of AVP as an ACTH-releasing hormone during insulin-induced hypoglycemia. To test the agent's capacity to inhibit an ACTH release by exogenous AVP, we compared the ACTH response to an infusion of 300 ng AVP/min a. 30 min after injection of 5 micrograms/kg of the antagonist, b. after injection of placebo (0.9% NaCl). Plasma ACTH levels during AVP infusion rose from 17.2 +/- 1.6 ng/l (3.8 +/- 0.35 pmol/l) to 31.7 +/- 4.2 ng/l (7.0 +/- 0.92 pmol/l) at 40 min after injection of the antagonist, the difference to the control-group (increment from 16.5 +/- 1.2 ng/l (3.6 +/- 0.26 pmol/l) to 41.8 +/- 3.5 ng/l) (9.2 +/- 0.77 pmol/l) being significant (p less than 0.05). Peak plasma cortisol levels were 323 +/- 42 and 529 +/- 52 nmol/l, respectively (p less than 0.05). We then tested the compound in the same subjects during an insulin-induced hypoglycemia; 30 min after administration of 10 micrograms/kg of the AVP antagonist or placebo, all subjects received 0.12 IU/kg of normal insulin, thus inducing a fall of blood glucose levels below 2 mmol/l. The AVP antagonist caused a moderate but insignificant reduction of the rise in plasma ACTH and a slightly greater, significant reduction of the increment in plasma cortisol (350 +/- 19 nmol/l with antagonist and 469 +/- 90 nmol/l with placebo, p less than 0.05) during insulin-induced hypoglycemia.(ABSTRACT TRUNCATED AT 250 WORDS)