BACKGROUND: Despite almost three centuries of use, the appropriate dosage of digitalis in patients with chronic heart failure and normal sinus rhythm has not been well studied. METHODS AND RESULTS: We studied 22 patients with heart failure who were receiving constant daily doses of digoxin, diuretics, and angiotensin-converting enzyme (ACE) inhibitors. In 18 patients, the oral daily dose of digoxin was increased from a mean of 0.20 +/- 0.07 to 0.39 +/- 0.11 mg/day corresponding to an increase in the serum digoxin concentration from 0.67 +/- 0.22 to 1.22 +/- 0.35 ng/mL. Radionuclide and echocardiographic left ventricular ejection fraction; maximal treadmill time; heart failure score; serum concentrations of norepinephrine, aldosterone, atrial natriuretic factor, and antidiuretic hormone; and plasma renin activity were obtained before and after the increase in digoxin dose. Subsequently, 9 patients were randomized to receive digoxin and 9 to receive placebo and radionuclide ejection fraction measured after 12 weeks. With the higher dose of digoxin compared with the lower dose, there was a significant increase in radionuclide ejection fraction from 23.7 +/- 9.6% to 27.1 +/- 11.8% (P = .007). No significant changes were noted in heart failure score; exercise tolerance; serum concentrations of norepinephrine, atrial natriuretic factor, and antidiuretic hormone; and plasma renin activity. There was, however, an increase in serum aldosterone concentration. Twelve weeks after the patients were randomized to receive digoxin or placebo, there was a significant decrease in ejection fraction (from 29.4 +/- 10.4% to 23.7 +/- 8.9%) in the placebo group but not in patients who continued to receive digoxin (P = .002). CONCLUSIONS: The increase in maintenance digoxin dose, while maintaining serum concentrations within therapeutic range, resulted in a significant increase in left ventricular ejection fraction that was not associated with significant changes in heart failure score, exercise tolerance, and neurohumoral profile.