To define the relationship between portal hypertension and renal excretion of salt and water, two acute animal models of portal hypertension were investigated. In both models, it was necessary for systemic and renal hemodynamics to remain unchanged during the creation of portal hypertension so as to eliminate the effects of change in these parameters on renal excretion.In eight dogs, portal hypertension was induced by controlled tightening of a ligature around the superior hepatic vein and changes in hemodynamics were prevented by controlled i.v. titration with canine plasma. Rises in portal pressure 9.3 k 2.6 to 15.6 f 2.3 cm HZO were associated with significant decreases in urine volume 1.57 k 0.53 to 0.66 k 0.21 ml per min (p < 0.05) and urinary sodium excretion 1340 & 82 to 145 k 76 mEq per min (p < 0.05)]. In contrast, similar experiments in five dogs in which the portal vein was partially ligated with increases in portal pressure from 8.3 f 0.6 to 18.6 & 0.5 cm HzO resulted in no change in urinary volume or sodium excretion.Acute reversible portal hypertension induced by hepatic vein constriction is associated with acute reversible retention of water and salt in the absence of changes in renal blood flow and creatinine clearance which only occur when the liver is involved and not with portal vein constriction.Many patients with advanced cirrhosis develop renal sodium retention and ascites. In the past, this sodium retention has been thought to be the result of fluid loss into the ascitic "third space" secondary to increased hydrostatic pressure in the hepatic sinusoids and splanchnic capillaries with a decrease in effective circulatory volume and appropriate alteration in renin and aldosterone secretion. Although plasma volume measurements in cirrhotics are actually normal or slightly elevated, it has been postulated that the ('central'' or nonsplanchnic component is decreased and the splanchnic component increased. Lieberman, in 1969, presented data from studies in human cirrhosis which suggested that renal sodium retention actually preceded the development of ascites, and was in fact, a contributing factor in the development of ascites (1). He postulated the existence of some hepatic factors which altered renal sodium handling independent of volume changes and the reninaldosterone regulatory mechanisms.