Comparatively high doses of thiazide diuretics have been implicated as a possible cause of ventricular arrhythmias. With the advent of lower recommended dosages, the effect of hydrochlorothiazide 50 mg daily alone and in combination with the potassium-sparing drug, amiloride 5 mg, on the frequency and severity of ventricular arrhythmias was examined in 37 elderly patients. The mean age was 81 +/- 2 years. The study used a randomized, double-blind, crossover design with 3 treatment phases: hydrochlorothiazide, hydrochlorothiazide + amiloride and placebo. A 24-hour Holter monitor recording and serum potassium measurement were obtained at the end of each treatment. Mean serum potassium was significantly (p less than 0.001) reduced with hydrochlorothiazide (3.5 +/- 1 mEq/liter) compared with placebo (4.1 +/- 0.1) and hydrochlorothiazide + amiloride (4.1 +/- 0.1). Serious ventricular arrhythmias occurred in 13 of 37 patients receiving hydrochlorothiazide compared with 15 of 37 receiving placebo and 9 of 37 receiving hydrochlorothiazide + amiloride. Patients who exhibited ventricular ectopy during the placebo phase tended to have an increase in the number of ventricular premature complexes receiving hydrochlorothiazide, although the difference was not statistically significant. There was a significant (p = 0.045) difference in the hourly ventricular premature complex frequency for hydrochlorothiazide compared with hydrochlorothiazide + amiloride. Patients with hypokalemia did not exhibit greater ventricular ectopic activity than those with a normal serum potassium, and concurrent digoxin therapy did not affect arrhythmia occurrence. Hydrochlorothiazide, 50 mg daily, did not affect the frequency or severity of ventricular ectopic activity in this elderly population.