Background: Data are limited regarding the patency of coronary artery bypass grafts performed by residents versus attending surgeons. Methods: We analyzed data from a multicenter, randomized Veterans Affairs Cooperative Study in which the left internal mammary artery was used preferentially to graft the left anterior descending coronary artery, and the best remaining coronary vessel received (per random assignment) either a radial artery or a saphenous vein graft. The study vessel's 1-year graft patency was the primary outcome measure. Secondary outcomes included operative times, operative morbidity, mortality, repeat revascularization, cost, angina symptoms, and quality of life. Multivariate analyses were used to compare patient outcomes for residents versus attendings. Results: Residents were designated as primary surgeons in 23% of cases (167 of 725). Among the 531 patients who had a 1-year angiogram, study graft patency rates for resident cases (n = 122) and attending cases (n = 409) were not significantly different (86% versus 90%, p = 0.22). Residents' cases had longer perfusion time (119 versus 105 minutes, p < 0.0001) and cross-clamp time (84 versus 68 minutes, p < 0.0001). After risk adjustment, all outcome measures did not differ between the two groups, and there was no apparent interaction effect between resident/attending designation and radial artery versus saphenous vein use or on-pump versus off-pump approach. Conclusions: Surgeons in training perform coronary artery bypass surgery without compromising graft patency or patient outcomes. Ongoing evaluation of residents' performance and surgical outcomes is needed, given the major changes that are occurring in residency training.