We examined the relationship between flow-mediated dilation (FMD) of the brachial artery and the extent and severity of coronary artery disease (CAD). Using high-resolution ultrasonography, we measured FMD and nitroglycerin-induced brachial artery dilation. We studied 121 patients (77 men, 44 women; mean age 63 +/- 11 years, range 25-79 years) who underwent coronary arteriography. The extent and severity of CAD were assessed by the coronary stenosis index (CSI). The adjusted FMD correlated inversely with CSI (r(s) = -0.63, p < 0.0001). Multivariate analysis demonstrated that the adjusted FMD was an independent predictor of CSI. The adjusted FMD was 10.2 +/- 4.8% in patients without CAD (n = 32), 7.7 +/- 6.0% in patients with single-vessel disease (n = 31), 5.2 +/- 5.5% in patients with double-vessel disease (n = 29), and 2.0 +/- 3.9% in patients with triple-vessel disease (n = 29). The adjusted FMD was significantly lower in the double- (p < 0.01) and triple-vessel (p < 0.0001) disease groups than in patients without CAD. The adjusted FMD was significantly lower in the triple-vessel disease group than in the single-vessel disease group (p < 0.001). Based on our results, as coronary atherosclerosis becomes more severe, the adjusted brachial artery FMD becomes more severely impaired.