Electrocardiographic evidence of left ventricular hypertrophy (ECG LVH), defined only by elevated R wave amplitudes in appropriate leads of the resting ECG, was ascertained among 2,760 survivors of myocardial infarction who had been randomized into the placebo treated group of the Coronary Drug Project. The entry ECG findings were related to the subsequent 3 yr mortality. Of 2,760 men, 186 (6.7%) had ECG LVH amplitude criteria in the baseline ECG when taken after at least 3 mth survival (average, 36 mth) from the last myocardial infarction. Three yr mortality in men with ECG LVH was almost twice that in men without ECG LVH (22.6% vs 12.1%; P = 0.0002) when the 'effect' of all other variables on mortality was ignored. However, when the relationship between baseline ECG LVH and subsequent mortality was adjusted, using a multivariate analysis, for other electrocardiographic and clinical findings, elevated R amplitude itself had no significantly independent prognostic importance. The prognostic importance of ECG LVH was largely explained by coexisting repolarization abnormalities, although ECG LVH plus abnormal ST segment and T wave findings had a less favorable prognosis than ST segment and T wave abnormalities alone. ECG LVH by R wave amplitude criteria in the absence of abnormal ST segment and T wave findings was unrelated to prognosis in these postinfarct patients.