OBJECTIVE To determine the effect of standard-dose trimethoprim-sulfamethoxazole on serum potassium concentration in hospitalized patients. METHODS Prospective chart review. METHODS Community-based teaching hospital. METHODS 105 patients with various infections were hospitalized and treated. Eighty patients treated with standard-dose trimethoprim-sulfamethoxazole (trimethoprim, < or = 320 mg/d; sulfamethoxazole, < or = 1600 mg/d) composed the treatment group; 25 patients treated with other antibiotic agents served as the control group. METHODS Serum sodium, potassium, and chloride concentrations; serum carbon dioxide content; anion gap; blood urea nitrogen level; and serum creatinine level. RESULTS The serum potassium concentration in the treatment group (mean +/- SD) was 3.89 +/- 0.46 mmol/L (95% CI, 3.79 to 3.99 mmol/L), and it increased by 1.21 mmol/L (CI, 1.09 to 1.32 mmol/L) 4.6 +/- 2.2 days after trimethoprim-sulfamethoxazole therapy was initiated. Blood urea nitrogen levels increased from 7.92 +/- 5.7 mmol/L (CI, 6.67 to 9.16 mmol/L) to 9.2 +/- 5.8 mmol/L (CI, 7.9 to 10.5 mmol/L), and serum creatinine levels increased from 102.5 +/- 49.5 mumol/L (CI, 91.4 to 113.6 mumol/L) to 126.1 +/- 70.7 mumol/L (CI, 110.3 to 141.9 mumol/L). Patients with a serum creatinine level of 106 mumol/L (1.2 mg/dL) or more developed a higher peak potassium concentration (5.37 +/- 0.59 mmol/L [CI, 5.15 to 5.59 mmol/L]) than patients with a serum creatinine level of less than 106 mumol/L (4.95 +/- 0.48 mmol/L [CI, 4.80 to 5.08 mmol/L]). Patients with diabetes had a slightly higher peak potassium concentration (5.14 +/- 0.45 mmol/L [CI, 4.93 to 5.39 mmol/L]) than did patients without diabetes (5.08 +/- 0.59 mmol/L [CI, 4.93 to 5.23 mmol/L]), but the difference was not statistically significant. The serum potassium concentration in the control group was 4.33 +/- 0.45 mmol/L (CI, 4.15 to 4.51 mmol/L), and it decreased nonsignificantly over 5 days of therapy. CONCLUSIONS Standard-dose trimethoprim-sulfamethoxazole therapy used to treat various infections leads to an increase in serum potassium concentration. A peak serum potassium concentration greater than 5.0 mmol/L developed in 62.5% of patients; severe hyperkalemia (peak serum potassium concentration > or = 5.5 mmol/L) occurred in 21.2% of patients. Patients treated with standard-dose trimethoprim-sulfamethoxazole should be monitored closely for the development of hyperkalemia, especially if they have concurrent renal insufficiency (serum creatinine level > or = 106 mumol/L).