Using external vascular ultrasound, we measured brachial artery diameter (Diam) at rest, after release of 4 min of limb ischemia, i. e., endothelium-dependent dilation (EDD), and after sublingual nitroglycerin, i.e., non-endothelium-dependent dilation (NonEDD), in 35 healthy men aged 61-83 yr: 12 endurance athletes (A) and 23 controls (C). As anticipated, treadmill exercise maximal oxygen consumption (VO(2 max)) was significantly higher in A than in C (40. 2 +/- 6.6 vs. 27.9 +/- 3.8 ml. kg(-1). min(-1); respectively, P < 0. 0001). With regard to arterial physiology, A had greater EDD (8.9 +/- 4.2 vs. 5.7 +/- 3.5%; P = 0.02) and a tendency for higher NonEDD (13.9 +/- 6.7 vs. 9.7 +/- 4.2%; P = 0.07) compared with C. By multiple linear regression analysis in the combined sample of older men, only baseline Diam (beta = -2.0, where beta is the regression coefficient; P = 0.005) and VO(2 max) (beta = 0.23; P = 0.003) were independent predictors of EDD; similarly, only Diam (beta = -4.0; P = 0.003) and VO(2 max) (beta = 0.27; P = 0.01) predicted NonEDD. Thus endurance-trained older men demonstrate both augmented EDD and NonEDD, consistent with a generalized enhanced vasodilator responsiveness, compared with their sedentary age peers.