One hundred consecutive patients undergoing aorta-coronary bypass grafting (ACBG) alone, without ventricular venting, were prospectively studied to determine the incidence and consequence of perioperative myocardial infarction (PMI) and the clinical variables that were predictive of PMI. Incidence was determined by serial electrocardiography (ECG) 100 patients; serum CK, GOT, and LDH (100 patients). CK isoenzymes (qualitative 100 patients, quantitated 50 patients); vectorcardiography (VCG) (78 patients); and 99mtechnetium pyrophosphate scintigraphy (TcPyp) (52 patients). The incidence of PMI by ECG was 9%; an additional 8% of cases was diagnosed by enzymes alone. The incidence of diagnostic change by VCG was 19% and by scintigraphy, 25%. Using at least one changed variable of the remaining three as the reference standard, the relative sensitivity and relative specificity of given variables in the diagnosis of PMI were as follows: ECG 67% and 100%, respectively; VCG 85% and 94%; scintigraphy 92% and 97%; and serum enzymes 86% and 96%. By univariate analysis, unstable angina was the only significant predictor of PMI. The operative mortality rate was 2% and the mortality rate at 12 months was 5%. There was a significantly greater mortality rate in patients with PMI diagnosed by ECG (p less than 0.01), in patients with unstable angina pectoris before operation (p less than 0.05), and in women (p less than 0.05).