To assess the 3 year risk of anterior myocardial infarction in patients with left anterior descending coronary artery territory disease (30 to 100% stenosis), National Heart, Lung, and Blood Institute (NHLBI) Coronary Artery Surgery Study (CASS) registry patients were identified who were 1) medically treated, and 2) had evidence of viable anterior myocardium at the time of baseline angiography. Prospectively, 118 patients having an anterior infarction within 3 years of baseline angiography were identified from annual follow-up of 4,535 medically treated patients who had left anterior descending coronary artery disease and viable anterior myocardium. From the large residual pool of patients without infarction, 141 were randomly selected from a stratified matrix to represent the entire group. The maximal percent stenosis was estimated by the CASS multiple angiographers, by a current single observer rereading and by contemporary computer measurement techniques. Absolute lumen dimension was assessed by computer measurement. The 3 year risk of anterior infarction was 2% for patients with their most severe left anterior descending stenosis less than 50%, 6% for patients with one such stenosis greater than or equal to 50% and 11% for patients with two or more such stenoses greater than or equal to 50% (p less than 0.02). Stenoses of 90 to 98% had the highest (15%) 3 year risk of anterior myocardial infarction. The three methods used to measure maximal percent stenosis differed little with regard to their predictiveness. Absolute lumen dimension was less predictive of risk. These results may provide a more rational basis on which to base coronary revascularization decisions.