Risk of clinically overt coronary heart disease in 190 persons with "definite" and 264 with "possible" electrocardiographic left ventricular hypertrophy (ECG-LVH) was compared with that of a total cohort of 5,127 men and women followed over 14 years. Prevalence of both coronary heart disease and ECG-LVH increased in proportion to antecedent blood pressure. Persons who acquired "definite" ECG-LVH had a residual threefold increased risk of clinically overt coronary heart disease after adjustment for the effect of coexisting hypertension. "Possible" ECG-LVH was associated with a twofold increased risk, but this was virtually obliterated on adjustment for hypertension. Risk of every clinical manifestation of coronary heart disease, and of death in particular, was increased, and 40% with prior ECG-LVH died in their initial attack, a fatality rate comparable with that of persons with prior overt coronary heart disease. ECG-LVH is thus a grave prognostic sign and a harbinger of clinically overt coronary heart disease. It is tempting to hypothesize that ECG-LVH without other explanation based mainly on increased voltage (possible LVH) is largely an expression of hypertensive hypertrophy and that with more marked voltage increases accompanied by S-T and T wave abnormality (definite LVH) indicates ischemic myocardial involvement.