OBJECTIVES: We sought to compare long-term outcomes after coronary bypass surgery with and without an internal thoracic artery graft. METHODS: We analyzed clinical outcomes over a median follow-up of 6.7 years among 3,087 patients who received coronary bypass surgery as participants in one of 8 clinical trials comparing surgical intervention with angioplasty. We used 2 statistical methods (covariate adjustment and propensity score matching) to adjust for the nonrandomized selection of internal thoracic artery grafts. RESULTS: Internal thoracic artery grafting was associated with lower mortality, with hazard ratios of 0.77 (confidence interval, 0.62-0.97; P = .02) for covariate adjustment and 0.77 (confidence interval, 0.57-1.05; P = .10) for propensity score matching. The composite end point of death or myocardial infarction was reduced to a similar extent, with hazard ratios of 0.83 (confidence interval, 0.69-1.00; P = .05) for covariate adjustment to 0.78 (confidence interval, 0.61-1.00; P = .05) for propensity score matching. There was a trend toward less angina at 1 year, with odds ratios of 0.81 (confidence interval, 0.61-1.09; P = .16) in the covariate-adjusted model and 0.81 (confidence interval, 0.55-1.19; P = .28) in the propensity score-adjusted model. CONCLUSIONS: Use of an internal thoracic artery graft during coronary bypass surgery seems to improve long-term clinical outcomes.