OBJECTIVE Recently published data from the North American Carotid Endarterectomy Trial revealed a benefit for carotid endarterectomy (CEA) in symptomatic patients with moderate (50% to 69%) carotid stenosis. This benefit was significant but small (absolute stroke risk reduction at 5 years, 6.5%; 22.2% vs 15.7%), and thus, the authors of this study were tentative in the recommendation of operation for these patients. To better elucidate whether CEA in symptomatic patients with moderate carotid stenosis is a proper allocation of societal resources, we examined the cost-effectiveness of this intervention. METHODS A decision-analytic Markov process model was constructed to determine the cost-effectiveness of CEA versus medical treatment for a hypothetical cohort of 66-year-old patients with moderate carotid stenosis. This model allowed the comparison of not only the immediate hospitalization but also the lifetime costs and benefits of these two strategies. Our measure of outcome was the cost-effectiveness ratio (CER), defined as the incremental lifetime cost per quality-adjusted life year saved. We assumed an operative stroke and death rate of 6.6% and a declining risk of ipsilateral stroke after the ischemic event with medical treatment (first year, 9.3%; second year, 4%; subsequent years, 3%). The hospitalization cost of CEA ($6,420) and the annual costs of major stroke ($26,880), minor stroke ($798), and aspirin therapy ($63) were estimated from a hospital cost accounting system and the literature. RESULTS CEA for moderate carotid stenosis increased the survival rate by 0.13 quality-adjusted life years as compared with medical treatment at an additional lifetime cost of $580. Thus, CEA was cost-effective with a CER of $4,462. Society is usually willing to pay for interventions with CERs of less than $60,000 (eg, CERs for coronary artery bypass grafting at $9,100 and for dialysis at $53,000). CEA was not cost-effective if the perioperative risk was greater than 11.3%, if the ipsilateral stroke rate associated with medical treatment at 1 year was reduced to 4.3%, if the age of the patient exceeded 83 years, or if the cost of CEA exceeded $13,200. CONCLUSIONS CEA in patients with symptomatic moderate carotid stenosis of 50% to 69% is cost-effective. Perioperative risk of stroke or death, medical and surgical stroke risk, cost of CEA, and age are important determinants of the cost-effectiveness of this intervention.