Stroke is a clinical diagnosis, but there is uncertainty as to the relative risks of surgery in patients with a visible infarction on brain CT. It is not clear whether visible infarction is associated with higher operative risk and poor outcome or is simply an indicator of a more "severe" stroke. Some studies have indicated that CT evidence of stroke carries an increased risk, other studies have found that patients with a low density and without significant shift on CT, with a stable neurological deficit and a normal level of consciousness can safely undergo carotid endarterectomy. Our aim was to answer a the specific question whether visible brain infarction on CT increases the operative risk of carotid endarterectomy, with special attention to the neurological complications. This analysis involved 105 patients with a visible brain infarct on CT scan who underwent carotid endarterectomy in Vilnius University Emergency Hospital. Carotid endarterectomy was performed in 86 men and in 19 women with mean carotid stenosis 87%, ranging from 42% to 99% by angiography. Clinical symptoms of stroke were present in 95 patients (86 carotid and 9 vertebrobasilar). Carotid endarterectomy underwent 76 patients with symptomatic carotid stenosis, 16 patients with symptoms of contralateral carotid artery, 9 patients with vertebrobasilar symptoms and 4 patients with nonspecific symptoms. Among symptomatic patients there were 46 patients with unilateral carotid stenosis and 30 patients with bilateral stenosis. Angiography of asymptomatic patients (n=29) revealed unilateral carotid lesions in 9 cases, bilateral stenosis in 8 patients and concomitant bilateral carotid and vertebral stenosis in 12 patients. The postoperative mortality rate was 2.9%, a new stroke occurred in 1.9%. The total rate of complications--4.8%. Patients with visible brain infarction on CT scan can safely undergo carotid endarterectomy with a standard operative risk, but risks modelling will be required if we are to target surgery most effectively at individuals.