Intracoronary infusions of acetylcholine (ACh) and the nitric oxide synthase (NOS) inhibitor NG -monomethyl-l-arginine (l-NMMA) are routinely used to assess endothelial function in the human coronary circulation. Currently, there are no studies that examine whether the response to one of these agents predicts the response to the other in a given individual. We sought to determine whether human coronary vasomotor responses to intracoronary ACh are predictive of the response to intracoronary l-NMMA. Following diagnostic coronary angiography, a coronary sinus catheter was placed. Sequential intracoronary infusions of ACh and l-NMMA were performed in 14 patients without severe coronary artery disease (CAD). In nine patients with severe CAD, only l-NMMA was infused. Quantitative coronary angiography was performed immediately following infusions to determine changes in coronary diameter. Coronary sinus oxygen saturation and the transcardiac oxygen saturation gradients were determined during infusions as an index of coronary blood flow (CBF). There were no significant differences in the prevalence of hypertension, diabetes, smoking or plasma creatinine or cholesterol levels between patients with and without severe CAD. In the 14 patients without severe CAD, ACh reduced coronary diameter on average by 8% (from 2.7 ± 0.8 to 2.5 ± 0.8 mm, P = 0.003) and reduced the transcardiac oxygen saturation gradient from 63 ± 8 to 42 ± 16% (P = 0.0001). N G -Monomethyl-l-arginine had no significant effect on coronary diameter (from 2.7 ± 0.8 to 2.7 ± 0.8 mm, n.s.) and did not alter the transcardiac oxygen saturation gradient (from 63 ± 8 to 63 ± 7%, n.s.). A positive rather than negative correlation was observed between the AChand l-NMMA-induced changes in coronary diameter (r = 0.752, P = 0.0002). In nine patients with severe CAD who did not receive ACh, l-NMMA had no effect on coronary diameter but produced an increased transcardiac oxygen saturation gradient consistent with reduced CBF (from 58 ± 5 to 62 ± 4%, P = 0.003). The CBF response to l-NMMA was significantly different between the two groups (P = 0.002). The failure of l-NMMA to reduce CBF following ACh infusion, in contrast to the results seen in the l-NMMA-only group, suggests that NOS had been rendered dysfunctional by the ACh infusion itself. The positive correlation in changes to coronary diameter between ACh and l-NMMA suggests that the degree of NOS dysfunction was directly correlated to the degree of preceding activation by ACh. Combined, these unexpected results suggest that ACh administration may acutely modify the function of NOS in the human coronary circulation.