Studies in our laboratory and others have demonstrated that arginine vasopressin (AVP) exerts potent vasoconstrictor actions on the vessels supplying the renal medulla. The physiological importance of these vascular effects of AVP has been difficult to assess because of high endogenous levels of AVP in anesthetized, surgically prepared animals. We have developed a decerebrated, hypophysectomized, renal-denervated rat model that enables us to study the effects of low levels of AVP on the pressure-diuresis, relationship under acute conditions. These rats maintain normal mean arterial pressure (MAP) and plasma AVP (2.5 pg/ml). Cortical and medullary blood flow (CBF and MBF, respectively) were measured by laser-Doppler flowmetry and total renal blood flow (RBF) by transit time flowmetry. Renal interstitial fluid pressure (RIFP) and urinary sodium excretion (UNaV) responses were determined during controlled increases of MAP produced by aortic occlusion below the renal arteries. From a baseline of 97 +/- 2 mmHg, 30% increases in MAP resulted in a 63% increase in MBF, 35% increase in RIFP, and sixfold increase in UNaV, whereas CBF and RBF remained unchanged. Infusion of AVP (0.50 ng.kg-1.min-1, which increased plasma AVP from normal control levels of 3 pg/ml to 11 pg/ml) produced no change in baseline MAP, RBF, or CBF but lowered MBF by 24%, RIFP by 26%, and UNaV by 71%. The slope of the relationship of AP and UNaV, MBF, and RIP was reduced to nearly zero by these small increases of plasma AVP. We conclude that an increase of plasma AVP in the range that occurs with water restriction decreases MBF selectively and greatly attenuates the arterial pressure-MBF and pressure-natriuretic relationship.