SUMMARY A review of 6040 consecutive exercise tests yielded 106 patients without previous myocardial infarction (MI) who had exercise-induced ST elevation (. 0.5 mm in a 15-lead ECG system). In 46, ST elevation was correlated with left ventriculography and coronary angiography. Coronary artery disease (CAD) (> 70% narrowing) was detected in 40 of 46 patients: 12 patients had one-vessel disease, 13 had two-vessel disease, and 15 had three-vessel disease. Resting ventriculograms were normal in 36 of 40 patients. Of 21 patients with anterior (V,-V3) ST glutamic oxaloacetic transaminase). All patients who did not fulfill the above criteria (i.e., those in whom an infarction was uncertain) were rejected from the study; patients who either fulfilled or had none of these criteria were included in the study.Of the 6040 patients tested, 399 (6.6%) had exerciseinduced ST elevation and a myocardial infarction. The majority (95%) of the 399 patients had ST elevation over the area of infarction. One hundred six patients (1.7% of the total population) developed ST elevation and had no evidence of a myocardial infarction, LVH or bundle branch block and were not taking digitalis. Forty-six patients in this latter group had undergone diagnostic cardiac catheterization. Another 21 patients with LVH and 12 patients taking digitalis also had exercise-related ST elevation. Six patients with LVH, one of whom was also taking digitalis, were catheterized. The stress tests were carried out according to either the multistaged Bruce or Balke protocols.'2 13 Each patient's blood pressure was monitored by indirect sphygmomanometry. Also, all patients were questioned during and after the exercise test about the development of symptoms. Angina pectoris was recorded if a patient developed exertional or postexer-