We examined by medical-record review why long-term digitalis therapy was prescribed in 150 outpatients, the reasons were: supraventricular tachyarrhythmias (35): supraventricular tachyarrhythmias and heart failure (33); and heart failure with sinus rhythm (82). In the patients without supraventricular tachyarrhythmias we scrutinized the diagnosis of heart failure using a clinicoradiographic scoring system and found the diagnosis unlikely in 32 patients. When these 32 patients are combined with the 31 patients who had only one occurrence of supraventricular tachyarrhythmias or heart failure, 42% of the patients were on long-term digitalis therapy for a questionable reason. We conclude that a substantial fraction of general medical outpatients might benefit from digitalis withdrawal, if evidence for heart failure is lacking or if the reason prompting digitalis therapy is isolated to the distant past.