The ability of quantitative thallium-201 scintigraphy to predict the extent and location of coronary artery disease before hospital discharge after acute myocardial infarction was evaluated in 52 patients. All patients underwent coronary angiography and serial thallium-201 imaging either at rest (10 patients) or after submaximal exercise stress (42 patients; target heart rate 120 beats/min). Two or three vessel disease was designated if abnormal thallium-201 uptake or washout patterns, or both, were seen in two or three vascular segments, respectively. Of 156 vessels analyzed in the 52 patients, 91 stenoses of 70 percent or greater were found by angiography. Seventy-four (81 percent) of these were predicted by scintigraphy. The specificity of scintigraphy for identifying vessel stenoses was 92 percent. Sensitivity for detecting and localizing stenoses supplying an infarct zone was 96 percent compared with 62 percent for stenoses supplying myocardium remote from the acute infarct. Perfusion abnormalities were more frequently seen in the distribution of vessels with severe (90 percent or greater) stenoses than in those with moderate (70 to 90 percent) stenoses (87 versus 53 percent, p less than 0.01). Scintigraphy detected a greater proportion of left anterior descending and right coronary arterial stenoses than circumflex stenoses (91 and 87 versus 63 percent, respectively, p less than 0.006). In the 42 patients who underwent submaximal exercise testing, multivariate analysis of 23 clinical and laboratory variables identified multiple thallium-201 defects as the best predictor of multivessel disease. The predictive accuracy of exercise-induced S-T segment depression was only 45 percent compared with 88 percent (p less than 0.05) for thallium-201 scintigraphy. Thus, 2 weeks after myocardial infarction, exercise thallium-201 scintigraphy is useful for predicting the extent and location of coronary artery disease, particularly stenoses in the left anterior descending and right coronary arteries. Moreover, thallium-201 imaging at rest is reliable in assessing the extent of coronary disease in hospitalized patients who cannot undergo exercise testing because of unstable angina, uncompensated heart failure, poorly controlled arrhythmias or physical limitations.