Abstract--Endothelin 1 mediates coronary vasoconstriction and endothelial dysfunction via endothelin receptor type A (ET_A) activation. However, the effects of selective endothelin receptor type B (ET_B) and combined ET_A+B receptor blockade on coronary vasomotion are unknown. We measured coronary vascular tone and endothelium-dependent and -independent vasomotor function before and after selective infusion of BQ-788 (an ET_B receptor antagonist) or combined infusion of BQ-788+BQ-123 (an ET_A antagonist) into unobstructed coronary arteries of 39 patients with coronary atherosclerosis or risk factors undergoing cardiac catheterization. BQ-788 did not affect epicardial diameter but constricted the microcirculation (P<0.0001), increased coronary sinus endothelin, and reduced nitrogen oxide levels. In contrast, BQ-123+BQ-788 dilated epicardial (P<0.0001) and resistance (P=0.022) arteries. Responses to acetylcholine and sodium nitroprusside were unaffected by BQ-788 alone. Epicardial endothelial dysfunction improved after BQ-123+BQ-788 (P=0.007). Coronary microvascular responses to acetylcholine and sodium nitroprusside were unaffected by BQ-123+BQ-788. We conclude that selective ET_B receptor antagonism causes coronary microvascular constriction, without affecting epicardial tone or endothelial function, via reduced endothelin clearance and NO availability. Combined ET_A+B blockade dilates coronary conduit and resistance vessels and improves endothelial dysfunction of the epicardial coronary arteries. Thus, endogenous endothelin, predominantly via ET_A receptor stimulation, contributes to basal constrictor tone and endothelial dysfunction, whereas ET_B activation mediates vasodilation in human coronaries. Our data suggest that selective ET_A blockade may have greater therapeutic potential than nonselective agents, particularly for treatment of endothelial dysfunction in atherosclerosis.