Objective: Patients with atherosclerotic aortic arch aneurysms are at greater risk for brain complication. We report our techniques and results of operation using selective cerebral perfusion. Methods: We retrospectively analyzed 52 consecutive patients with atherosclerotic aortic arch aneurysms (mean age, 70 years, range, 53±86 years), who underwent operation between April 1992 and March 1999. The operation was non-elective in 11 patients (21.1%). Concomitant operations included eight coronary artery bypass grafting and one aortic valve replacement. Simultaneous distal aortic reconstruction was performed in three patients. The operation was performed through median sternotomy. To avoid brain embolism, total arch replacement with a branched prosthesis was performed in 48 patients, in an attempt to exclude affected segments of aorta. In addition, retrograde femoral artery perfusion was avoided and cerebral circulation was isolated before aortic manipulation. To achieve even blood¯ow distribution, we employed perfusion and continuous pressure monitoring of all the three arch vessels. The perfusion rate was 12^2 ml/kg per min and the pressure was kept around 50 mmHg. Deep hypothermic arrest of the lower torso (bladder temperature, 228C) was used during open distal aortic anastomosis. Results: The hospital mortality rate was 11.5% (six of 52), and 7.3% (three of 41) for elective cases. Only one patient (1.9%) developed permanent focal neurological de®cit. Six other patients showed temporary brain complications, which was global (delirium) in three and focal in three others. Conclusions: Selective cerebral perfusion is a safe brain protection method, and our strategy seems effective for embolic stroke prevention. q