This study aimed to clarify the impact of carotid or intracranial atherosclerosis on perioperative stroke in patients undergoing open aortic arch surgery.Between 2008 and 2015, 200 consecutive patients underwent elective aortic arch surgery with selective antegrade cerebral perfusion and moderate hypothermic circulatory arrest. Nonselective screening for carotid or intracranial atherosclerosis was performed using carotid ultrasonography or magnetic resonance angiography. Carotid or intracranial atherosclerosis was classified as below moderate (0%-49% stenosis), moderate (50%-69%), or severe (70%-100%). In patients with moderate or severe stenosis, cerebral hemodynamics were evaluated using single-photon emission computed tomography with acetazolamide.None of the 37 patients undergoing preoperative single-photon emission computed tomography with acetazolamide had impaired cerebral hemodynamics. In-hospital mortality rate was 3.5% (7/200). Postoperative neurologic morbidity included permanent stroke in 8 patients (4.0%) and transient neurologic deficits in 27 patients (14%). Permanent stroke occurred in 3 of 159 patients (1.9%) with below moderate stenosis and 5 of 41 patients (12.2%) with moderate or severe stenosis (P = .008). Seven of 8 patients (87.5%) with stroke experienced multiple atherothrombotic embolizations, and 1 patient experienced a stroke of unknown cause. In multivariate analysis, previous cerebrovascular accident (odds ratio, 5.0; 95% confidence interval, 2.07-12.42; P = .0004) and shaggy aorta (odds ratio, 4.2; 95% confidence interval, 1.58-10.98; P = .0045) were significant determinants of neurologic morbidity.Embolism was the major cause of permanent stroke in our patient population. Preoperative craniocervical and aortic screening may aid in modifying the operative strategy to reduce the occurrence of stroke.