BACKGROUND The North American Symptomatic Carotid Endarterectomy Trial (NASCET) and the Asymptomatic Carotid Atherosclerosis Study (ACAS) showed significant risk reductions for carotid endarterectomy (CE) but did not consider the cost-effectiveness of CE. RESULTS We developed Markov models based on NASCET and ACAS to simulate hypothetical cohorts of patients with carotid stenosis and calculated quality-adjusted life expectancies and direct medical costs for those receiving either CE or medical therapy. For symptomatic patients, we used a surgical stroke risk of 5.8%, a 2-year stroke risk of 27.6% for medical patients, and a post-30-day surgical risk reduction of 87% for ipsilateral strokes. For symptom-free patients, we used a surgical stroke risk of 1.7%, a 5-year stroke risk of 17.5% for medical patients, and a post-30-day surgical risk reduction of 74% for ipsilateral strokes. Cost and quality-of-life estimates were estimated from the literature. The incremental cost-effectiveness ratio of CE versus medical therapy was $4100 and $52700 per quality-adjusted life year (QALY) gained for symptomatic and symptom-free patients, respectively. Incremental cost effectiveness ratios were < $50000/QALY gained for symptomatic patients for wide variations in baseline assumptions. For asymptomatic patients, the incremental cost-effectiveness ratio was $100900/QALY gained if the perioperative risk was 4%, $36400/QALY gained if the risk of untreated patients was doubled, and $13500/QALY gained if the cost of CE was halved. CONCLUSIONS Performance of CE is associated with favorable incremental cost-effectiveness ratios compared with other accepted medical interventions; however, the analysis for symptom-free patients was sensitive to a number of parameters.