Objective: The purpose of this study was to evaluate the effect of target vessel characteristics on radial artery patency when used as a composite T graft. Methods: Between October 1993 and March 2001, 1022 patients underwent coronary bypass with the internal thoracic artery-radial artery composite T graft. Of these, angiography has been performed on 109 patients at a mean 27.1 months (range, 2-70 months) postoperatively for symptoms of ischemia. By means of Cox proportional hazard models, the relationships between anastomotic patency and target vessel location, proximal stenosis, target diameter, and quality were assessed. Results: A total of 231 radial artery anastomoses were evaluated. The mean stenosis for patent anastomoses was 82% compared with 71% for occluded anastomoses (P < .001). Anastomotic patency for targets with moderate stenosis (≥70%) was worse than that for vessels with critical stenosis (≥90%; relative risk, 1.7; 95% confidence interval, 1.3-2.2; P < .001). Patency for targets of the right coronary artery was statistically inferior to that for targets of the left anterior descending artery (relative risk, 1.8; 95% confidence interval, 1.2-2.9; P = .01) and bordered on significance versus that for the circumflex artery distribution (relative risk, 1.6; 95% confidence interval, 1.0-2.8; P = .06). When directed toward critically stenosed targets of the left anterior descending or circumflex arteries., radial artery patency was not statistically different that that of the internal thoracic artery (P =.19). Conclusion: Radial artery patency is sensitive to both target location and proximal target stenosis. Selective use of the radial artery to targets of the left anterior descending and circumflex distributions remains encouraging. Radial artery grafts to targets of the right coronary artery or those with moderate stenosis tenosis appear to be at particularly high risk of failure.