Previous studies have suggested that the early post-infarction exercise test is useful in predicting the extent of coronary artery disease. The results of a heart rate limited exercise test three weeks after infarction and a symptom limited exercise test six weeks after infarction obtained by both standard lead electrocardiograms and 16 lead precordial maps were compared in 100 consecutive survivors of acute myocardial infarction under 55 years of age. Exercise tests were defined as being positive on the basis of angina, ST segment depression greater than or equal to 1 mm in any electrocardiogram lead, or exertional hypotension. Multivessel disease, that is two or three vessel disease, was present in 60 patients, and three vessel disease in 22 patients. The sensitivity, specificity, and predictive value for multivessel disease of the three week test were 38%, 83%, and 76% respectively; and results for the six week test were 55%, 75%, and 77% respectively. Only 32% of patients with three vessel disease were identified at the three week test, and 59% at the six week test. Significantly more patients with multivessel and three vessel disease were identified by the symptom limited six week test. Precordial mapping offered no advantages over the standard 12 lead electrocardiogram in either the identification of patients with multivessel disease or the prediction of the distribution of coronary artery disease. Angina pectoris during the exercise test at six weeks was the single most useful predictor of multivessel disease. Multivessel disease was found in 27 (87%) of the 31 patients with angina with or without ST depression during the test at six weeks compared with 33 (48%) of the 69 patients who did not have angina during the test at six weeks. Exercise testing in the early post-infarction period in patients under 55 years of age is of limited value in predicting the extent of coronary artery disease. It is, therefore, unreasonable to use such exercise tests to select patients for coronary arteriography after myocardial infarction. None the less angina pectoris occurring during a symptom limited exercise test six weeks after infarction is a strong predictor of multivessel disease, and coronary arteriography is recommended in these patients.