The nonfiltering kidney model was used to determine whether sodium or potassium inhibits renin secretion in the absence of a functional macula densa in dogs with thoracic inferior vena caval constriction. The control rate of renin secretion was high, and decreases were readily recognized. After control observations, hypertonic sodium chloride or hypertonic potassium chloride was infused into the renal artery for 1 hour, and renin secretion was measured at 15-minute intervals. An increase in renal venous plasma sodium concentration from 141 to 154-158 mEq/liter caused no change in renin secretion for the first 45 minutes of infusion in dogs with a nonfiltering kidney. In contrast, dogs with thoracic caval constriction but with a filtering kidney showed a striking decrease in renin secretion during intrarenal sodium infusion (3,097 to 1,061 ng angiotensin/min, P <0.02). An increase in renal venous plasma potassium concentration from 3.9 to 6.1 mEq/liter in one group of dogs and from 4.9 to 8.3 mEq/liter in a second group also caused no change in renin secretion in the nonfiltering kidney of dogs with thoracic caval constriction. However, in dogs with thoracic caval constriction and a filtering kidney, potassium infusion decreased renin secretion (1,952 to 984 ng angiotensin/min, P <0.05). In all experiments, infusion of sodium chloride or potassium chloride failed to produce a significant change in renal blood flow, arterial blood pressure, or inferior vena caval pressure. Therefore, no evidence was obtained for a vascular action or a direct effect of sodium or potassium on the juxtaglomerular cells, and the data are consistent with an action mediated by the renal tubular system.