Nineteen men with mild to moderate hypertension and without a history of cardiac arrhythmias were randomized (double-blind) into groups to receive hydrochlorothiazide (HCTZ) at a dose of 25 mg/day, HCTZ at 50 mg/day, or HCTZ (25 mg) plus triamterene (50 mg) for a six-month period after a three-week (single-blind) placebo period. Serum electrolyte values were determined at baseline and at frequent intervals thereafter. Twenty-four hour ambulatory electrocardiograms were obtained at baseline and just before study termination. Mild hypokalemia (less than 3.5 mEq/L) occurred in approximately half of the patients and was unrelated to treatment group. Serious arrhythmias were infrequent, though some patients had large numbers of extra beats. The incidence of arrhythmia appeared unrelated to serum potassium concentration. We conclude that mild hypokalemia associated with low-dose diuretic therapy for hypertension is not arrhythmogenic.