The cross-sectional area (CSA) of large-conductance arteries increases in response to endurance training in humans. To determine whether training-induced changes in arterial structure are systemic in nature or, rather, are confined to the arteries supplying exercising muscles, we studied 10 young men who performed one-legged cycle training [80% of one-legged peak O2 uptake (VO2 peak)), 40 min/day, 4 days/wk] for 6 wk and detraining for another 6 wk. There were no significant differences in baseline one-legged VO2 peak) and CSA of the common femoral artery and vein (via B-mode ultrasound) between experimental and control legs. In the experimental leg, one-legged VO2 peak) increased 16% [from 3.0 +/- 0.1 to 3.4 +/- 0.1 (SE) l/min], arterial CSA increased 16% (from 84 +/- 3 to 97 +/- 5 mm2), and venous CSA increased 46% (from 56 +/- 5 to 82 +/- 5 mm2) after endurance training. These changes returned to baseline during detraining. There were no changes in one-legged VO2 peak) and arterial CSA in the control leg, whereas femoral venous CSA in the control leg significantly increased 24% (from 54 +/- 5 to 67 +/- 4 mm2) during training. Changes in femoral arterial and venous CSA in the experimental leg were positively and significantly related to corresponding changes in one-legged VO2 peak) (r = 0.86 and 0.76, respectively), whereas there were no such relations in the control leg (r = 0.10 and 0.17). When stepwise regression analysis was performed, a primary determinant of change in VO2 peak) was change in femoral arterial CSA, explaining approximately 70% of the variability. These results support the hypothesis that the regional increase in blood flow, rather than systemic factors, is associated with the training-induced arterial expansion. Femoral arterial expansion may contribute, at least in part, to improvement in efficiency of blood transport from the heart to exercising muscles and may facilitate achievement of aerobic work capacity.