Increased plasma osmolality results in increased central as well as peripheral release of vasopressin. Experiments were carried out to determine whether, in this circumstance, vasopressin can act centrally to modulate its peripheral release. Prior to the start of a thirty-min i.v. infusion of 2.5 M or 0.15 M NaCl, the rats were given an intracerebroventricular (i.c.v.) injection of a peptide V1/V2 vasopressin antagonist (2 micrograms), OPC-31260 (60 micrograms), a non-peptide V2 antagonist, or 1-desamino-8-D-arginine vasopressin (dDAVP, 5 ng), a V2 agonist. Experiments with the peptide antagonist were carried out in male and non-estrous female rats. Since there were no differences between males and females in the measured responses, experiments with the other two drugs were carried out only in males. Pretreatment with either the V1/V2 antagonist or the V2 antagonist enhanced the increase in plasma vasopressin levels in response to the hypertonic saline infusion by about 50% at the end of 30 min. dDAVP, on the other hand, had no effect. None of the i.c.v. drugs had an affect on either the pressor or bradycardic responses to hypertonic saline infusion. These observations suggest that vasopressin can act centrally in a negative feedback fashion to attenuate its own release into the peripheral circulation in response to increased plasma osmolality.