AbstractRetrograde perfusion via the coronary sinus supplies vascular beds distal to coronary stenoses and has been used for administration of cardioplegia. An additional application is to supply noncardioplegic retrograde perfusion while performing proximal anastomoses (a time when cardiac arrest is not critical). The aim of this study was to determine the safety of this technique and to study the metabolic changes with antegrade versus retrograde warm blood perfusion. Sixty‐six patients, with good left ventricular function, underwent distal coronary bypass in a similar fashion. Proximal anastomoses were done with 1) partial occlusion clamp (n = 29) or 2) cross‐clamp on and continuous, warm, noncardioplegic retrograde blood perfusion (n = 37). In an additional 10 patients, metabolism was assessed with antegrade and retrograde perfusion during proximal anastomoses. Despite longer cross‐clamp times (96.4 ± 6.2 vs 80.8 ± 3.1 min, p<0.05) with retrograde perfusion, the total duration of cardiopulmonary bypass was significantly less (119.6 ± 6.2 vs 136.6 ± 4.6 min, p<0.05). There was superior postbypass, intraoperative hemodynamics (cardiac index) with retrograde perfusion (4.0 ± 0.2 vs 3.6 ± 0.1 L/min/m2). The incidence of postoperative dysrhythmia was not significantly different between groups. Oxygen and glucose utilization was more efficient with retrograde perfusion. Retrograde perfusion during proximal anastomoses is a safe technique. There is diminished risk of aortic dissection, atheroembolism, delayed aneurysm formation, or rupture due to avoidance of application of partial occlusion clamps. There is evidence of superior substrate