Both retrograde perfusion via the coronary sinus and direct perfusion of the coronary artery during circulatory occlusion under selective brain cooling were studied and evaluated from the electrocardiographic standpoint. The following results were obtained. 1) Hypothermia is not applicable for this type of experiment because irritability of myocardium is increased by hypothermia resulting ventricular fibrillation. 2) Direct perfusion is more physiologic than retrograde perfusion and circulatory cessation is maintained over 50 minutes without deleterious effect to myocardium compared to only 15 minutes available in the latter method. 3) Perfusion pressure not exceeding 35 mm Hg. in retrograde perfusion and below 120 mm Hg. in direct perfusion are optimal to obviate mycardial damage which is manifested as ST-T deviation in electrecardiogram. It is safe to keep temperature of perfusion blood between 36 and 37°C. Blood containing over 7.2 gl/dl Hb is applicable for perfusion if oxygenation is satisfactorily attained.