To examine the role of mineralocorticoid and glucocorticoid in potassium (K) tolerance in healthy humans, we studied the effects of canrenoate, a mineralocorticoid antagonist, and RU486, a glucocorticoid antagonist, on the excretion of a KCl load. Canrenoate (200 mg, iv) or RU486 (400 mg, orally) was administered 150 min before a KCl load (1 mmol/kg BW, orally) in seven healthy males undergoing maximal water diuresis. Clearance studies were extended for 5 h after the KCl load, and the data were compared with time control, KCl load alone, and canrenoate alone. KCl increased K excretion (from 18.8 +/- 2.4 to 63.3 +/- 3.9 mmol/5 h; P < 0.01) and sodium (Na) excretion (from 35.9 +/- 2.1 to 72.9 +/- 6.0 mmol/5 h; P < 0.01). Clearance calculations, based on maximal water diuresis, were compatible with increased distal Na and volume delivery. Canrenoate alone modestly increased basal cumulative NaCl excretion and had no effect on K excretion. However, canrenoate blunted the kaliuresis after the KCl load (51.9 +/- 4.4 mmol/5 h; P < 0.05 compared to KCl alone) and stimulated natriuresis in a complementary way. Clearance data were compatible with diminished distal Na reabsorption and K secretion in response to an undisturbed KCl-induced increase in distal Na delivery. RU486 did not influence the excretion of the KCl load or its effects on renal sodium handling parameters, although effective glucocorticoid receptor blockade was likely to be present in view of the increase in plasma cortisol. These data suggest that in healthy humans, mineralocorticoid activity, but not glucocorticoid activity, is involved in the elimination of a K load. The latter contrasts with data in adrenalectomized animals, in which situation glucocorticoid as well as aldosterone are indispensible for normal K tolerance.