In choosing a pharmacologic agent for stress testing, the clinician must keep a number of things in mind, such as the diagnostic utility of the agent or in what situations a vasodilator or catecholamine will be the better choice. Although all stress agents produce similar diagnostic accuracy for CAD, vasodilators have a higher cardiac uptake than catecholamines, and the addition of exercise improves the heart/background contrast ratios. With regard to physiologic comparisons, exercise or dobutamine will double coronary perfusion compared with baseline flow, but vasodilators produce a threefold or fourfold increase. The clinician should also keep in mind that adenosine will produce the shortest duration of hyperemia, whereas dobutamine and arbutamine produce a longer effect, and dipyridamole has the longest duration. If electrophysiologic considerations are important, exercise and catecholamines accelerate sinoatrial and atrioventricular conduction and are not typically associated with heart block. In contrast, adenosine can cause transient atrioventricular block, but this rarely occurs with dipyridamole. Clinical factors also must be considered. Although clinical utility of pharmacologic stress agents in the first 24 hours after infarction has not been demonstrated, the prognostic utility of vasodilators in the subsequent 2- to 4-day period has been shown. With patients with pulmonary disease (asthma) who do not have wheezing, dipyridamole can be used, but dobutamine or arbutamine should be used in patients with recent respiratory failure or bronchospasm before testing. In patients with left bundle branch block, vasodilators are the preferred stress agents rather than synthetic catecholamines or dynamic exercise. In the first crossover thallium imaging, there was good overall agreement in segmental perfusion comparing adenosine and dipyridamole, but there was a tendency for adenosine to detect more ischemia. The clinical significance (if any) for these findings has yet to be determined.