As a general rule, clinicians should attempt to isolate those patient subsets whose posttest patient management may be optimally changed by the test referral. Use of testing in a heterogeneous population will lead to an increase in the false-positive rate with a resultant increase in the rate of normal catheterization and higher costs of care. By providing imprecise test results to the clinician, a higher rate of cardiac events may be observed as a result of lack of care for these patients. In addition, a high rate of normal coronary angiograms in patients with positive test results or admissions for subsequent myocardial infarctions in patients with normal test results may result in a lack of confidence in the imaging results leading to a change in referral to newer, untested modalities. Thus to integrate our knowledge of the benefits to exercise electrocardiography within the context of clinical decision making and the appropriate selection of a noninvasive test, it is clear that physicians may use aggregate indexes such as the Duke treadmill score to guide patient care, but they should not be relied on to provide definitive evidence. Nor can one make the statement that because of the risk stratification ability of the Duke treadmill score this modality may be used to supplant stress myocardial perfusion imaging. In fact, exercise electrocardiography should be the test of choice for lower risk patient populations. However, in more intermediate-risk populations (e.g., known coronary disease or abnormal resting electrocardiogram), stress perfusion imaging is of established diagnostic and prognostic value. Further, from the exercise treadmill test, patients with an intermediate Duke treadmill score would benefit from additional noninvasive testing with stress myocardial perfusion imaging. This strategy provides substantial prognostic information on which to guide subsequent patient care.