In a prospective study the question was investigated whether the exercise ECG, in patients who had sustained a transmural myocardial infarction two to 36 months previously, provides information on perfusion of the ventricular myocardium (apart from infarct area); in other words, whether the coronary system can be assessed beyond the area of the "infarct vessel". Results obtained with the exercise ECG and coronary angiography were correlated for 81 patients with typical ECG sign of a previous anterior-wall (35) or posterior-wall infarct (46). In patients with an old posterior-wall infarct multiple-vessel disease was predicted with a 79.2% sensitivity, since ST segment depression in V3-V6 gave evidence of hemodynamically significant stenoses (stenosis of greater than 50% of vessel diameter). There was a 72.7% specificity. In patients with an old anterior-wall infarct any additional stenosis was predicted with a 78.6% sensitivity and 76.2% specificity. However, the high rate of multiple-vessel disease in symptomatic patients (63.0% after posterior-wall infarct, 62.5% after anterior-wall infarct), as well as the serious consequences of failure to miss potentially operable multiple-vessel disease in this at-risk group patients, reduce the value of this method for indicating the need of coronary angiography. In asymptomatic patients (frequency of multiple-vessel disease 36.8% after posterior-wall infarct, 21.1% after anterior-wall infarct), on the other hand, exercise ECG (predictive value of ST depression 71.4% after posterior-wall infarct, 50% after anterior-wall infarct) provides reasonable additional help in deciding whether coronary angiography is essential in at-risk patients.