To test the hypothesis that the release of neurohypophyseal peptides into plasma in humans is stimulated by a central nervous system (CNS) alpha 1 adrenergic mechanism, we measured the responses of arginine vasopressin (AVP) and oxytocin (OT) to intravenous methoxamine, an alpha 1 agonist which enters the CNS following peripheral administration. The potential confound of baroreceptor inhibition of AVP release by the pressor effect of methoxamine was addressed by measuring the plasma AVP response to infusion of norepinephrine (NE), an alpha 1 agonist which does not enter the CNS and which produced an equivalent pressor effect. We also assessed the pituitary adrenocortical system responses to methoxamine and norepinephrine infusions by measuring plasma ACTH and cortisol concentrations. In addition, plasma NE and epinephrine were measured. Methoxamine, but not NE, increased plasma AVP compared to placebo infusion. Neither methoxamine nor NE affected plasma OT. The AVP elevation was delayed until more than 60 min after the methoxamine infusion began and the peak AVP level occurred 30 min after cessation of the infusion. In contrast, ACTH and cortisol increased early during methoxamine infusion and ACTH returned to baseline promptly after the infusion ceased. Although it is possible that the AVP response to methoxamine reflected stimulation of AVP release at a CNS level, it is also possible that the AVP increase represented a rebound response to withdrawal of methoxamine.