Indwelling catheters were placed in conscious dogs into the vena portae via a jejunal mesenteric vein several weeks prior to actual experiments. To compare a possible different response to different pathways of administration, 5% saline solution was infused either into vena portae or into vena cava at a rate of 0.05 ml/min/kg BW for 40 min. 1. Infusion of 5% sodium chloride into the vena portae, as well as into the vena cava caused a significant increase in urine flow, serum osmolality, sodium excretion rate and heat rate. The increase in urine flow was always due to an increase in osmolal clearance whereas free water clearance decreased consistantly. In some experiments plasma renin activity (PRA) was measured by radioimmunoassay. For both routes of administration the hypertonic saline infusion caused a decrease in PRA. At no time did any of these parameters display significant differences depending on route of application. 2. This means that the calculated differences between infused and excreted amounts of water and of sodium did not depend on the route of infusions. 3. A close relationship was found between initial urine osmolality and the length of the periods during which peak values of urine flow were observed.