The study objective was to determine the impact on outcome associated with using the second internal thoracic artery as a free compared with an in situ graft among patients who received the first internal thoracic artery as an in situ internal thoracic artery to the left anterior descending artery. Among 2600 patients who underwent bilateral internal thoracic artery with an in situ internal thoracic artery to the left anterior descending artery, the second internal thoracic artery was used as a free graft bilateral internal thoracic artery in 136 patients and as an in situ graft (in situ bilateral internal thoracic artery) in 2464 patients. One-to-many propensity score matching was performed to produce a cohort of 134 patients with a second free graft internal thoracic artery matched to 2359 patients with a second in situ internal thoracic artery. Early and long-term outcomes including survival, hospital readmission, and repeat revascularization up to a maximum of 25.8 years were compared. There were no differences between the 2 matched groups' preoperative baseline characteristics and early adverse events. Long-term survival at 5, 10, and 15 years was significantly higher among patients with an in situ bilateral internal thoracic artery compared with patients with a free graft bilateral internal thoracic artery (hazard ratio free graft bilateral internal thoracic artery vs in situ bilateral internal thoracic artery, 1.53; 95% confidence interval, 1.14-2.10; P = .004). However, the long-term risk of readmission to the hospital for cardiovascular reasons and need for repeat revascularization were not significantly different between the 2 matched groups. In a small, propensity-matched cohort of patients undergoing coronary artery bypass grafting, the use of a second in situ internal thoracic artery was associated with an increase in late survival compared with the use of a second internal thoracic artery as a free graft. However, the risk of late hospital readmission and the need for repeat revascularization were similar. Display omitted