This study was performed to evaluate the efficacy of various treatment modalities for hyperkalemia in 8 end-stage renal disease (ESRD) patients. Simultaneous administration of sodium bicarbonate and insulin with glucose was compared with infusion of either bicarbonate alone or insulin and glucose. Plasma potassium was measured at the baseline and after 60 min of infusion with each regimen. Infusion of 8.4% solution of sodium bicarbonate at 2 mEq/ min for 60 min induced a significant rise in blood bicarbonate from 21.7 ± 2.1 to 26.3 ± 1.7 mEq/l (p < 0.01), but failed to lower plasma potassium (6.4 ± 0.1 vs. 6.3 ± 0.2 mEq/l, before and after). Intravenous infusion of insulin and glucose (5 mU/kg/min for 60 min) significantly lowered plasma potassium from 6.3 ± 0.1 to 5.7 ± 0.1 mEq/l (p < 0.01). The combined infusion of bicarbonate and insulin with glucose showed the greatest decline in plasma potassium, from 6.2 ± 0.2 to 5.2 ± 0.1 mEq/l (p < 0.01). With the combined regimen, the increases in plasma bicarbonate (22.3 ± 1.7 to 25.8 ± 1.9 mEq/l, p < 0.05) and blood pH (7.36 ± 0.02 to7.42 ± 0.02, p < 0.01) were significant, but somewhat less than those with bicarbonate administration alone. Plasma insulin levels before treatment were similar in all treatment regimens, and increased markedly following the infusion of insulin with glucose, either with or without sodium bicarbonate (9 ± 1.5 vs. 10 ± 10 μU/ml before insulin, and 196 ± 18.0 vs. 201 ± 26.4 μU/ml after insulin). Plasma epinephrine, norepinephrine, osmolality and plasma aldosterone before and after treatment did not show any significant differences among the 3 different regimens. In conclusion, the ineffectiveness of sodium bicarbonate alone and its synergistic effect with insulin and glucose in acute therapy of hyperkalemia in ESRD patients suggest that mild metabolic acidosis, which is common in patients on maintenance hemodialysis, may contribute to tissue insensitivity to the action of insulin on transcellular potassium shift.