Objective: To compare the outcome of surgical high-risk elderly patients with severe aortic stenosis(SAS) treated by different therapy procedures, including transcatheter aortic valve implantation(TAVI), surgical aortic valve replacement(SAVR), and drug therapy. Methods: We retrospectively analyzed the clinical data of 242 surgical high-risk elderly (age ≥65 years old) SAS patients hospitalized in Fuwai Hospital between September 2012 and June 2015. According to the treatment method, patients were divided into TAVI group (81 cases), SAVR group (59 cases) and drug therapy group (102 cases). The primary end point was all-cause mortality at 1 year post procedure, and secondary end point included cardiac function class(NYHA), vascular complication, valvular function, non-fatal myocardial infarction, new atrial fibrillation, stroke, bleeding, pacemaker implantation, acute renal failure, and readmission. We used the Kaplan-Meier method to estimate survival function based on follow up data and survival was compared between groups with the use of the log-rank test. Results: (1) In the baseline data, there were statistically significant difference among 3 groups for the age, left ventricular ejection fraction, cardiac function class Ⅲ and Ⅳ, rates of combined diabetes, chronic renal failure, mild and moderate mitral regurgitation (P<0.01 or 0.05). The risk score of the Society of Thoracic Surgeons(STS) was 7.28±4.98 in the TAVI group, and 5.67±3.49 in the SAVR group(P=0.036). (2) The perioperative rates of pacemaker implantation(11.3%(9/81) vs. 0, P=0.025) and mild paravalvular regurgitation(29.6%(24/81) vs.1.7%(1/59), P<0.001) were significantly higher in TAVI group than in SAVR group.(3)The rate of rehospitalization was significantly lower in TAVI group than in SAVR group(3.0%(2/67) vs. 22.7%(10/44) P=0.005) and the rate of pacemaker implantation was significantly higher in TAVI group than in SAVR group(17.5 (12/67) vs. 0, P=0.008) after 1 year. The rates of death from any cause in the TAVI (5.8%(4/67)) and SAVR group (11.4%(5/44)) were significantly lower than that in the drug therapy group (54.9%(50/91), both P<0.05) after 1 year and was similar between TAVI group and SAVR group(P=0.622). (4) The rates of cardiac function classⅠandⅡ increased and Ⅲ and Ⅳ decreased in TAVI and SAVR group after 1 year when compared with base line(P<0.001). The rates of cardiac function class Ⅱ, and Ⅲ increased , class Ⅰ and Ⅳ decreased in drug therapy group after 1 year compared with base line (P=0.020). (5)The survival rates after 1 year were significantly higher in the TAVI group and SAVR group than in the drug therapy group(log-rank test, P<0.001), and the difference was similar between TAVI group and SAVR group (log-rank test, P=0.062). Conclusion: In surgical high-risk elderly patients with SAS, the prognosis of drug therapy was poor, and TAVI and SAVR were associated with similarly improved rates of survival after 1 year, although there were significant differences in periprocedural complications between TAVI and SAVR groups.