The chronic effects of potassium loading on sodium balance and related variables were studied in two groups of dogs. The first group was intact except for the presence of indwelling arterial and venous cannulas. On the 1st day, increasing daily potassium intake from a normal level (30 mEq/day) to 200 mEq/day produced a 0.47-mEq increase in plasma potassium with a 56% increase (P greater than 0.01) in sodium excretion in spite of a 58% increase in plasma aldosterone concentration. After 6 days of potassium loading the cumulative negative sodium balance averaged 44 mEq while 22Na space decreased 6.7% (P less than 0.025). In this group arterial pressure did not change measurably. The same experimental protocol was repeated in a second group of dogs that were chronically adrenalectomized and maintained on fixed levels of aldosterone (50 mug/day) and hydrocortisone (1 mg/day). With aldosterone levels held constant the same increase in potassium intake produced a 1st day increase in potassium concentration of 1.20 mEq/liter and 217% (P less than 0.001) increase in sodium excretion. After 5 days of high potassium intake, the cumulative negative sodium balance totaled 84 mEq. Sodium space decreased 7.5% (P less than .005) during the course of the 5-day high potassium intake period. Potassium loading caused a fall in mean arterial pressure in this group; pressure fell from the control level of 110 +/- mm Hg to 87 +/- 4 mm Hg (P less than .001) after 3 days of high intake. By the 5th day of the experiment, pressure stabilized at 96 +/- 3 mm Hg, 13% less than (P less than 0.01) the control level. The results suggest that changes in plasma potassium concentration within physiological limits may have long term effects on sodium balance.