Patients having hyperkalemia often are given bicarbonate to raise blood pH and shift extracellular potassium into cells. Blood pH in many hyperkalemic patients, however, is compensated. To determine whether bicarbonate, independent of its pH action, affects plasma potassium, 14 hyperkalemic patients were treated with bicarbonate in 5% dextrose. In five patients (changed pH group), blood pH rose at least 0.08, while in nine (constant pH group), it changed less than 0.04. In the first group, pH rose 0.12, bicarbonate rose 5.9 mEq/liter, and plasma potassium fell 1.6 mEq/liter, and plasma potassium fell 1.4 mEq/liter. The correlation between changes in plasma potassium and bicarbonate was identical in the two groups and independent of urinary potassium excretion. Four additional patients, who were treated with 5% dextrose alone, did not significantly lower their plasma potassium, although subsequent treatment with bicarbonate in 5% dextrose lowered their plasma potassium. Thus, bicarbonate lowers plasma potassium, independent of its effect on blood pH, and despite a risk of volume overload, should be used to treat hyperkalemia in compensated acid-base disorders, even in the presence of renal failure, provided the plasma bicarbonate concentration is decreased.