Seventy-seven patients underwent aortic arch aneurysm repair using selective cerebral perfusion from January 1987 to August 1992. Early and long-term results and preoperative and postoperative cerebral function were evaluated. Cerebral function was assessed by the mini mental state-Himeji test and the Wechsler adult intelligence scale. Thirty-six patients had true aneurysms, and 41 had dissection. Hospital mortality for true and dissecting aneurysms was 19.4% and 7.3%, respectively. The 5-year actuarial survival rates for true and dissecting aneurysms were 59.0% and 65.3%, respectively (not significant). There were no significant differences in test scores before or after operation. Repair or replacement of the aortic arch using selective cerebral perfusion is a safe procedure with acceptable hospital mortality.