It is uncertain whether, in humans, potassium depletion can cause or sustain metabolic alkalosis of clinically important degree in the absence of coexisting known alkalosis-producing conditions. Previously we found, in normal humans ingesting abundant NaCl, that dietary K+ depletion alone can induce and sustain a small decrease in blood acidity and increase in plasma bicarbonate concentration; we hypothesized that more severe alkalosis was prevented by mitigating mechanisms initiated by renal retention of dietary NaCl that was induced by K+ depletion. To ascertain the acid-base response to dietary K+ depletion under conditions in which the availability of NaCl for retention is greatly limited, in the present study of six normal men we restricted dietary K+ as in the previous study except that intake of NaCl was maintained low (2 to 7 mEq/day, Low NaCl Group) instead of high (126 mEq/day, High NaCl Group). Plasma acid-base composition and renal net-acid excretion (NAE) did not differ significantly between groups during the control period. In the steady state of K+ depletion (days 11 to 15 of K+ restriction), neither plasma K+ concentration (2.9 +/- 0.9 mEq/liter vs. 3.0 +/- 0.1 mEq/liter) nor cumulative K+ deficit (399 +/- 59 mEq vs. 466 +/- 48 mEq) differed significantly between groups. During K+ restriction, persisting metabolic alkalosis developed in both groups, which was more severe in the Low NaCl Group: increment in [HCO3-]p, 7.5 +/- 1.0 mEq/liter versus 2.0 +/- 0.3 mEq/liter, P less than 0.001; decrement in [H+]p, 5.5 +/- 0.6 nEq/liter versus 2.9 +/- 0.4 nEq/liter, P less than 0.003. A significantly more severe alkalosis in the Low NaCl Group was evident at all degrees of K+ deficiency achieved during the course of the 15 days of K+ restriction, and the severity of alkalosis in the Low NaCl Group correlated with the degree of K+ deficiency. During the generation of alkalosis (days 1 to 7 of K+ restriction), NAE increased in the Low NaCl Group whereas it decreased in the High NaCl Group. During the maintenance of alkalosis (days 11 to 15), NAE stabilized in both groups after it returned to values approximating the control values. In both groups, urine Cl- excretion decreased during K+ restriction even though Cl- intake had not been changed, with the result that body Cl- content increased negligibly in the Low NaCl Group (28 +/- 6 mEq) and substantially in the High NaCl Group (355 +/- 64 mEq).(ABSTRACT TRUNCATED AT 400 WORDS)