Left-sided coronary revascularization with bilateral internal thoracic artery (BITA) graft is performed usually either with an in situ (double source) or Y-graft configuration (single source). Two hundred fifty-three (mean age, 67.1 ± 9.5 years) patients underwent isolated left-sided coronary revascularization with BITA graft alone at the present authors’ institution (2000–2015). Skeletonized BITA grafts were used either in an in situ (n = 199) or Y-graft configuration (n = 54). Forty pairs were identified with the propensity score-matching. Outcomes of the two groups were compared both in unmatched and matched series. Cardiopulmonary exercise testing was performed in five pairs of selected, asymptomatic matched patients having patent BITA grafts at coronary computed tomography angiography. BITA in situ patients had lower risk profiles than BITA Y-graft patients (median EuroSCORE II, 1.9 vs. 2.9%, p = 0.051). In-hospital mortality (5.6 vs. 0, p = 0.0093) and the rates of postoperative complications except deep sternal wound infection were higher in BITA Y-graft patients. However, these differences were not confirmed in matched groups. During the follow-up period (mean, 5.9 ± 4.3 years), between BITA in situ and BITA Y-graft matched patients, there were no differences in non-parametric estimates of freedom from cardiac death (p = 0.6), major adverse cardiac and cerebrovascular events (MACCEs, p = 0.65), and repeat coronary revascularization (p = 0.44). Adjusted risk estimates of MACCEs according to BITA configuration confirmed no superiority of the one configuration over the other (p ≥ 0.44). No significant differences were found at the cardiopulmonary exercise testing. Results of left-sided coronary revascularization with BITA graft alone are independent from BITA configuration, even after stress testing.