The dynamics of potassium excretion were examined in normal dogs and dogs with chronic renal insufficiency of at least 4 weeks' duration (remnant model). All animals, in balance on diets providing 15, 50, or 100 mEq of potassium and 100 mEq of sodium, were challenged with 50 mEq of potassium chloride. Immediately thereafter, hourly clearances were obtained for 5 hours. Irrespective of dietary potassium, mean fasting serum potassium and urinary potassium excretion (UKV) were similar in normal and remnant dogs with mean GFR's of 57 +/- 3 and 16 +/- 3 ml/min, respectively. After orogastric administration of 50 mEq potassium, serum potassium rose significantly more in remnant (2.2 to 2.5 mEq/liter) than in normal (0.9 to 1.2 mEq/liter) groups (P less than 0.001). Conversely, UKV increased significantly less, 70 to 96 vs. 151 to 194 micro Eq/min, respectively (P less than 0.001). In 5 hours, normal animals excreted 61 to 67% of the load, but remnant dogs only 30 to 37% (P less than 0.001). In all groups, UKV correlated directly with serum potassium concentration. But this relationship was markedly attenuated in the remnant groups (P less than 0.001) and independent of dietary potassium. In contrast, the same slope describes the relationship between UKV/GRF and serum potassium for all, normal and remnant, animals. The blunted kaliuresis occurred despite the more severe hyperkalemia in remnant than in normal dogs; it was not associated with significant changes in acid-base, diuresis, natriuresis, serum glucose, insulin, and glucagon concentrations and occurred despite prolonged hyperaldosteronism. The results demonstrate a severe limitation of the remnant kidney's ability to rapidly excrete a potassium load. Changes in serum potassium, or a consequence thereof, are important for the urinary excretion of potassium following its acute administration.