Coronary vasomotion has an important role in the regulation of myocardial perfusion. During dynamic exercise, normal coronary arteries dilate, whereas stenotic arteries constrict. This exercise-induced vasoconstriction has been associated with the occurrence of myocardial ischemia and has been believed to be the result of endothelial dysfunction, with a reduced release or production of EDRF, increased sympathetic stimulation, enhanced platelet aggregation with release of thromboxane A2 and serotonin, or a passive collapse of the disease-free wall segment within the stenosis (the Bernoulli effect), or a combination of any of these. More recently, it has been realized that pharmacological treatment might prevent exercise-induced vasoconstriction and, thus, reduce myocardial ischemia and the occurrence of angina pectoris. Vasodilators such as nitrates, calcium antagonists or alpha-receptor blockers dilate the coronary arteries and prevent coronary stenosis narrowing during exercise. In contrast, beta-blocking agents are associated with coronary vasoconstriction at rest, but--conversely--can induce coronary vasodilatation during exercise. Pharmacological treatment in patients with stable angina pectoris may improve myocardial ischemia by reducing pre- and afterload, myocardial contractility, oxygen consumption, and vasomotor tone. However, coronary collateral perfusion can modify these effects by shunting blood from the non-ischemic to the ischemic region (collateral flow) or by shunting blood from the ischemic to the non-ischemic zone (coronary steal phenomenon). Typically, a steal phenomenon has been reported in patients receiving either dipyridamole or calcium antagonists, whereas a reversed steal has been described after beta-blockade, with an increase in contralateral tone shunting blood from the non-ischemic to the ischemic zone (reverse steal phenomenon).