Serum albumin concentration has been strongly associated with risk of death in hemodialysis patients, with mortality increasing as albumin decreases. Metabolic acidosis stimulates protein catabolism and decreases protein synthesis. A study was undertaken to investigate the effect of increasing predialysis serum bicarbonate (HCO3) concentrations on the nutrition of hemodialysis patients as measured by albumin and total lymphocyte count (TLC). Metabolic acidosis was defined as a predialysis serum bicarbonate concentration of < or = 18 mEq/L. Thirty-six hemodialysis patients were enrolled in the study. Each had been stable on hemodialysis for > or = 3 months and each had a mean serum bicarbonate concentration of < or = 18 mEq/L on predialysis monthly laboratory values during the preceding 3 months. The subjects were randomized into 2 groups. The first group consisted of 18 control subjects who were dialyzed on a standard bicarbonate bath of 35 mEq/L. The second group consisted of 18 experimental patients who were dialyzed on a bicarbonate bath of 40 mEq/L. Subjects in the experimental group who had predialysis serum bicarbonate concentrations less than 22 mEq/L after 2 weeks on the higher bicarbonate bath were additionally supplemented with oral sodium bicarbonate at a dosage of 1 mEq/kg dry weight/d. Monthly predialysis laboratory values were checked for all subjects and included serum electrolytes, blood urea nitrogen, calcium, and albumin. TLCs were obtained at the initiation and at the conclusion of the study. Intact parathyroid hormone, blood pressures, and interdialytic weight gains were also followed. The study lasted 16 weeks; 32 subjects completed the study (16 in each group). There were no statistically significant differences between the two groups at the initiation of the study. The serum bicarbonate concentrations were significantly different between the two groups at the end of the study (control HCO3 17.3 +/- 3.2 mEq/L v experimental HCO3 20.2 +/- 2.9 mEq/L; P = 0.01). Serum albumin concentrations and TLCs were not statistically different (P > 0.05) between the two groups at the end of the study (control albumin 3.88 +/- 0.28 g/dL v experimental albumin 3.76 +/- 0.26 g/dL and control TLC 1,780.0 +/- 779.4/mm3 v experimental TLC 2,020.1 +/- 888.0/mm3). Potassium, intact parathyroid hormone, interdialytic weight gain, blood pressures, Kt/Vs, and protein catabolic rates did not differ. We found that the change in serum bicarbonate concentration was well-tolerated and was without any demonstrable side effects. We conclude that increasing the serum bicarbonate concentration by 3 mEq/L for 16 weeks has no effect on the indicators of nutrition that we measured (serum albumin and TLC).