To assess the risks and potential benefits of coronary artery bypass graft surgery (CABG) after myocardial infarction (MI), we retrospectively evaluated 129 patients admitted within 48 hours of MI who subsequently underwent diagnostic coronary and left ventricular angiography within 60 days of MI. Surgical candidates were objectively defined as patients who had CABG after angiography or patients who did not have CABG but had left ventricular ejection fraction > 25% and two or more residual, jeopardized ventriculographic segments. Ninety-four of the 129 patients (73%) were surgical candidates: 43 underwent CABG and 51 were managed medically. Seventy of these 94 patients (74%) had no angina, 86 (91%) had no heart failure symptoms immediately preceding angiography. Surgical patients had a significantly greater number of coronary arteries with ≥ 70% stenosis (2.5 ± 0.1 vs 1.9 ± 0.1, p = 0.0001), an operative mortality of 2% (one of 43) and a perioperative infarction rate of 5% (two of 43). At a mean follow-up time of 23 ± 2 months after MI, cardiac mortality was 7.0% (three of 43) in the surgical group and 16% (eight of 51) in the medically managed surgical candidates (p = NS). However, patients with four to 10 residual, jeopardized ventriculographic segments who were managed surgically had significantly improved survival compared with those managed medically (93 ± 5% vs 64 ± 11%, p < 0.05). The frequency of sudden death during the follow-up period was 0% (zero of 43) in surgical patients and 12% (six of 51) in medically managed surgical candidates (p < 0.05). Thus, within the limits of this retrospective study, CABG after MI may constitute reasonable management for patients who have extensive residual myocardial jeopardy and may improve longevity by preventing sudden death.