OBJECTIVE Bilateral internal thoracic artery (BITA) grafting in patients with diabetes are controversial because of increased risk of sternal infection. On the other hand, patients with diabetes may benefit from BITA grafts because of the associated improved survival. This study evaluated factors affecting early and long-term outcomes for better selection of patients with diabetes for BITA grafts. METHODS Between 1996 and 2006, 69 patients with insulin-treated diabetes and 732 with orally treated diabetes received isolated skeletonized BITA grafts. Of these patients, 338 were younger than 65 years, 322 were between 65 and 74 years old, and 141 were 75 years or older. RESULTS Operative mortality was lower than logistic EuroSCORE-calculated mortality (2.9% vs 7%, P < .001). Predictors of increased mortality were critical preoperative state (P < .001) and age (P = .008). There were 30 cases of sternal infection (3.7%); predictors were reoperation (P < .001), peripheral vascular disease (P = .009), obesity (P = .012), chronic lung disease (P = .009), and female sex (P = .020). Mean follow-up was 8.4 ± 4 years. Kaplan-Meier 10-year survivals were 75%, 59%, and 39% for patients younger than 65, 65 to 74, and at least 75 years, respectively (P < .001). They were better than corresponding Charlson comorbidity index-predicted survivals (36%, 10%, and 3%, respectively; P < .001). Predictors of decreased survival were age (P < .001), congestive heart failure (P < .001), and peripheral vascular disease (P < .001). Off-pump surgery was independently associated with better long-term survival (P = .003). CONCLUSIONS BITA grafts are safe in patients with diabetes. Favorable short- and long-term outcomes outweigh increased sternal infection risk.