The prognostic utility for predicting cardiac events was determined for dipyridamole-thallium scintigraphy, exercise stress testing (when possible; n = 69) and multiple clinical variables in 100 consecutive patients admitted for elective surgical repair of peripheral vascular disease. After initial noninvasive evaluation, 11 patients were referred for coronary angiography and the remaining 89 patients had surgery without further cardiac studies. Fifteen patients (17%) had a postoperative myocardial infarction, one of which was fatal. Of these 15 patients, 14 had thallium redistribution and 3 had positive ST segment depression during stress testing. Among the many variables tested, the presence of redistribution on serial dipyridamole-thallium images was the most significant predictor of serious cardiac events. All 11 patients who had coronary angiography had both redistribution and multivessel coronary artery disease. Four of these 11 patients died during follow-up and 6 had coronary artery bypass surgery. It is concluded that dipyridamole-thallium imaging has significant prognostic utility in predicting postoperative myocardial infarction and death in patients with severe peripheral vascular disease, and is superior to exercise testing or clinical variables in determining cardiac risk. The odds for a serious cardiac event were 23 times greater in a patient with thallium redistribution than in a patient without redistribution, strongly suggesting that myocardial imaging may be used as a primary screening test before elective vascular surgery.