To assess long-term prognostic significance of transient ischemia in patients with documented coronary artery disease and stable symptoms and to examine the relation between transient ischemia and the site of angiographic disease progression following acute cardiac events. Cohort study with a mean+/-SD follow-up of 51.5+/-23.8 months. Ambulatory patients with stable coronary artery disease, assigned to medical therapy. A total 221 patients (173 men; mean age, 60.8 years) were recruited. Of the 221 patients, 101 (45.7%) had single-vessel, 86 (38.9%) had 2-vessel, and 34 (15.4%) had 3-vessel disease. A total of 135 had a positive exercise test for ischemia, and mean+/-SD resting left ventricular ejection fraction (LVEF) was 49.8%+/-11.4%. Using conventional criteria, patients were prospectively stratified as low risk for continued medical therapy (single-vessel disease, 2-vessel disease with negative exercise test, or LVEF> or =40%; n=189 [85.5%]) or high risk for continued medical therapy (multivessel disease with ischemia and/or left ventricular dysfunction; n=32 [14.5%]). Ambulatory ST-segment monitoring, treadmill exercise testing, radionuclide ventriculography, and coronary angiography. Demographic, clinical, ambulatory monitoring, treadmill exercise, and left ventricular function variables as independent predictors of acute (cardiac death, myocardial infarction, or unstable angina) or all (including revascularization) cardiac events in the overall and the low-risk population. None of the clinical or noninvasive measures of ischemia were of prognostic significance in the overall or the low-risk group. The only significant independent predictor of outcome in all patients for all events, including revascularization, was the number of diseased vessels (X2=13.5 [df=1]; P<.001). Exclusion of vessel disease resulted in conventional risk stratification as the most significant predictor of outcome from all events in all patients (X2= 10.3 [df= 1]; P=.001). In the low-risk group, the number of diseased vessels was the only predictor for all events (X2=4.6; P=.03). For acute cardiac events, none of the variables tested were of prognostic significance. Based on the frequency of events in the low-risk patients, a 2-fold increase in the rate of cardiac events in patients with transient ischemia compared with those without transient ischemia during ambulatory monitoring could be excluded with greater than 85% power and alpha of .05. Of 30 patients suffering acute nonfatal cardiac events during follow-up, angiography was performed in 27, revealing significant progression of coronary disease in 24 (88.8%) and the development of new significant lesions at sites remote from previously significant lesions in 20 (74%) cases. These new lesions were equally likely to occur in those with or without transient ischemia at initial assessment. Acute cardiac events in predominantly low-risk stable angina patients with confirmed coronary disease are unpredictable, and those more likely to suffer such an event cannot be identified by the detection of ambulatory ischemia. Acute nonfatal cardiac events result predominantly from the development of significant new coronary lesions, not initially severe enough to cause ischemia. Patients categorized as high risk for long-term medical therapy have an increased rate of cardiac events (mainly revascularization) when compared with low-risk patients.