The incidence of congestive heart failure (CHF), derived from more than 30 years of follow-up, is examined by electrocardiogram (ECG) and radiography in relation to cardiac hypertrophy. Cardiac failure occurred in 485 of 5,209 subjects participating in the Framingham Study. Hypertension was the dominant predisposing factor for both cardiac hypertrophy and cardiac failure. The ECG pattern of left ventricular hypertrophy (ECG-LVH) heralded serious cardiovascular disease of all varieties, but risk ratios were two- to fivefold greater for the development of CHF in men and women (ages 35-64 years) than for any other sequelae. Risk of CHF in those with ECG-LVH exceeded that for unrecognized ECG patterns at myocardial infarction (ECG-MI). The ECG pattern of left ventricular hypertrophy, characterized by increased voltage unaccompanied by a repolarization abnormality, carried a decreased risk, chiefly reflecting the severity of coexistent hypertension. The independent contribution of ECG-LVH with accompanying repolarization changes to the risk of CHF was equal in the two sexes and persisted with advancing age. The ECG pattern of left ventricular hypertrophy was more strongly associated with occurrence of CHF than was radiographic enlargement, and contributed to the risk of CHF (taking radiographic heart size into account). Echocardiographic evidence of LVH (ECHO-LVH) was more common in subjects with CHF than was ECG-LVH, occurring in 63% of women and 77% of men with CHF, and LVH was the most frequently observed echocardiographic finding. Cardiac hypertrophy was found to be an ominous harbinger of cardiac failure, particularly when it was manifested on an ECG with repolarization abnormality.(ABSTRACT TRUNCATED AT 250 WORDS)