In large epicardial coronary arteries, β2 adrenoceptors represent only 15% of the total population of β adrenoceptors.1 Thus, their stimulation is unlikely to affect coronary blood flow significantly.2 Conversely, in small coronary arteries, β2 adrenoceptors represent 93% of the total population of β adrenoceptors.1 Their stimulation causes vasodilation.2 Experimental studies, however, have shown that β2-adrenoceptor-mediated vasodilation of small coronary arteries is more pronounced in subepicardial than in subendocardial layers, thus provoking reduction of the endo/epicardial flow ratio.3 As the subendocardium is more susceptible to myocardial ischemia than the subepicardium,4 this redistribution of coronary blood flow toward the subepicardial layers, in the presence of critical coronary stenoses, might reduce the ischemic threshold during exercise. Furthermore, stimulation of β2 adrenoceptors in the sinus node, recently demonstrated in humans,5 might increase the heart rate, thus reducing exercise duration. We investigated the clinical importance of the detrimental consequences of β2-adrenoceptor stimulation in a trial of the effects of salbutamol on exercise-induced myocardial ischemia in patients with stable angina pectoris. The confounding influence of β1-adrenoceptor stimulation was prevented by pretreatment with atenolol. Furthermore, to establish whether nonselective β-adrenoceptor blockade is able to prevent these detrimental effects, salbutamol was given to another group of patients after pretreatment with propranolol. © 1991.