OBJECTIVES: The objectives were to compare the effectiveness and cost-effectivness of carotid artery stent with emboli protection device (CAS+EPD) and carotid endarterectomy (CEA). METHODS: A retrospective cohort during 2001-2012 was built using NHI database. Outcomes events cumulative incidence rate of death, stroke, death or stroke, and MI occurring 30 days peri-procedure, 1 year and 8 years after procedure, were analyzed as comparative effectiveness. The transitional probabilities of various outcomes were adopted from comparative effectiveness results by Weibull distribution. Kaplan Meier sample average method was applied for medical cost. A Markov model was built to simulate the lifetime QALYs and medical costs estimation. RESULTS: A total of 3,359 and 543 patients were included in CAS+EPD and CEA groups. In comparative effectiveness, the incidences of safety outcomes in stroke (2.2% vs. 2.0%), death (0.7% vs. 1.5%), and death or stroke (2.7% vs. 2.9%) did not differ significantly between CAS+EPD and CEA within 30 days post procedure. A one-year follow-up revealed that CEA was associated with higher risks of stroke (hazard ratio: 2.72, 95%CI: 1.61-4.61) and death or stroke (HR: 2.00, 95%CI: 1.33-3.02) than CAS+EPD. Long term follow-up results demonstrated CEA had a higher risk in stroke (HR: 1.61, 95%CI: 1.09-2.37) only. The hospitalization cost were $5,600±2,500 in CAS+EPD and $ 4,800±6,100 in CEA, the total medical expense during the first year were $11,600 and $10,000, respectively.  Life-long medical cost estimation revealed $28,700 for CAS+EPD and $31,300 for CEA. Cost-effectiveness analysis showed CAS+EPD had 0.59 life years (LYs) gained better than CEA (9.24 LYs vs. 8.65 LYs). The QALYs for CAS+EPD and CEA were 8.12 and 6.99, respectively. Overall, the results demonstrated CAS+EPD to be the dominant strategy. CONCLUSIONS: Retrospective cohort database analysis demonstrated CAS+EPD was more effective and also less expensive than CEA. Under current NHI reimbursement, the CAS+EPD was a cost-effective strategy.