1. Indwelling non-occlusive catheters were placed in the vena porta and inferior vena cava of female rats several days before experimentation. Isotonic saline or isosmotic glucose (2% of body wt.) was infused into one vein followed one to several days later with an identical infusion into the other vein of each conscious animal. 2. Significantly higher urine flow and sodium excretion resulted from infusion of isotonic saline (0.5 ml/min) into the vena porta than into the vena cava. Modest prehydration or section of the hepatic branch of the right vagus did not affect the differential sodium response. Changes in endogenous creatinine clearance and potassium excretion were not significantly different for the two routes. Mean values for net peak sodium excretion and contemporaneous urine flow, urinary sodium concentration, and fractional sodium excretion were significantly higher for the portal than for the caval infusion while differences in glomerular filtration rate and filtered sodium load were insignificant. No significant difference in sodium excretion resulted from infusion of isosmotic glucose by the two routes. 3. Compared to the response promoted by the isotonic saline load infused at 0.5 ml/min, the differential response in sodium excretion was prolonged when the same load was infused at 0.375 ml/min. Sodium excretion was not significantly different for the two routes when the same isotonic saline load was infused at 0.75 ml/min. 4. These experiments provide evidence for participation of the liver in the control of sodium excretion and suggest release of a hepatic humoral factor which may be controlled by the duration of exposure of the hepatic circulation to an adequate load of isotonic saline.