Marfan syndrome is an inherited disorder characterized by genetic abnormality of microfibrillar connective tissue proteins. Endothelial dysfunction is thought to cause aortic dilation in subjects with a bicuspid aortic valve; however, the role of endothelial dysfunction and endothelial damaging factors has not been elucidated in Marfan syndrome. Flow-mediated dilation, a noninvasive measurement of endothelial function, was evaluated in 39 patients with Marfan syndrome. Aortic diameter was measured at the aortic annulus, aortic root at the sinus of Valsalva, sinotubular junction and ascending aorta by echocardiography, and adjusted for body surface area (BSA). The mean value of flow-mediated dilation was 6.5 ± 2.4 %. Flow-mediated dilation had a negative correlation with the diameter of the ascending thoracic aorta (AscAd)/BSA (R = -0.39, p = 0.020) and multivariate analysis revealed that flow-mediated dilation was an independent factor predicting AscAd/BSA, whereas other segments of the aorta had no association. Furthermore, Brinkman index had a somewhat greater influence on flow-mediated dilation (R = -0.42, p = 0.008). Although subjects who smoked tended to have a larger AscAd compared with non-smokers (AscA/BSA: 17.3 ± 1.8 versus 15.2 ± 3.0 mm/m(2), p = 0.013), there was no significant change in flow-mediated dilation, suggesting that smoking might affect aortic dilation via an independent pathway. Common atherogenic risks, such as impairment of flow-mediated dilation and smoking status, affected aortic dilation in subjects with Marfan syndrome.