Current management of coronary artery disease has taken a very aggressive approach in which cardiac catheterization plays a prominent role in patient evaluation and in which bypass and angioplasty are commonly used for therapy. The number of cardiac catheterizations and procedures, not surprisingly, have grown in tandem because angiography provides anatomic information that becomes the substrate for justifying interventions. Bypass surgery has been shown to confer a survival benefit compared with medical therapy in patients with multiple-vessel disease and left ventricular dysfunction, but it also is still used in other patient populations with equivocal indications. Comparison studies of percutaneous transluminal coronary angioplasty with medical management have indicated a slight advantage with percutaneous transluminal coronary angioplasty in limiting symptoms, but no evidence yet supports its survival benefit. Angioplasty, however, has become much more common in the last decade, particularly as the initial revascularization technique. Because cardiac catheterization is frequently the nodal branch point between invasive and noninvasive (i.e., medical) management, its application should be limited to high-risk candidates who would receive a survival benefit from these procedures or to those with intractable symptoms. Those who propose that catheterization is the best method for risk stratification argue that noninvasive testing requires physiologically significant disease and that morbid or fatal events can occur with rapid progression of minimal disease. From the studies reviewed, however, the extent of coronary angiographic disease is not clearly predictive of future cardiovascular events. Although some studies found the number of diseased vessels to be independent prognostic variables, most found that it did not add any additional prognostic information beyond that provided from noninvasive testing. Furthermore, there has been an argument that silent ischemia puts patients at higher risk of sudden death or infarction, and these patients need to be catheterized. However, numerous studies have shown that this concern is exaggerated. The studies reviewed found that except for patients with diabetes, those with "silent" or painless exercise-induced ST depression do not have a worse prognosis than those with symptomatic ST depression when cardiovascular death, sudden death, or acute myocardial infarction are considered Clinical and exercise test variables have been underused in estimating prognosis. Specifically, they are rarely used systematically to stratify patients into low-risk groups who do not need catheterization and high-risk groups who should undergo angiography because of possible lesions amenable to bypass or angioplasty.(ABSTRACT TRUNCATED AT 400 WORDS)