The view that digitalis clinically benefits patients with heart failure and sinus rhythm lacks support from a well-controlled study. Using a randomized, double-blind, crossover protocol, we compared the effects of oral digoxin and placebo on the clinical courses of 25 outpatients without atrial fibrillation. According to a clinicoradiographic scoring system, the severity of heart failure was reduced by digoxin in 14 patients; in nine of these 14, improvement was confirmed by repeated trials (five patients) or right-heart catheterization (four patients). The other 11 patients had no detectable improvement from digoxin. Patients who responded to digoxin had more chronic and more severe heart failure, greater left ventricular dilation and ejection-fraction depression, and a third heart sound. Multivariate analysis showed that the third heart sound was the strongest correlate of the response to digoxin (P less than 0.0001). These data suggest that long-term digoxin therapy is clinically beneficial in patients with heart failure unaccompanied by atrial fibrillation whose failure persists despite diuretic treatment and who have a third heart sound.