BACKGROUND The study sought to assess the clinical profile, outcome, and predictors for mortality of "real-world" high-risk severe aortic stenosis patients according to the mode of treatment assigned. METHODS Patients were referred to a dedicated clinic for meticulous screening and multidisciplinary team assessment and 343 were finally assigned treatment (age 81.3 ± 7.2 years, 42.3% men): transcatheter aortic valve replacement (TAVR) with the Edwards SAPIEN or CoreValve device, 100 (29.2%); surgical aortic valve replacement (SAVR), 61 (17.8%); balloon valvuloplasty (as definitive therapy), 27 (7.9%); medication only, 155 (45.2%). No patient was lost to follow-up. RESULTS The balloon valvuloplasty group had a significantly higher 1-month mortality rate (18.5%) than the TAVR group (3%, P = 0.006) and medical therapy group (3.9%; P = 0.004), without significant difference from the SAVR group (11.5%, P = 0.5). One-year cumulative survival was significantly higher in the TAVR group (92%) than in the other groups (SAVR 71%, balloon valvuloplasty 61.5%, medication 65%; all P < 0.001). Among survivors, 1-year rates of high functional class (NYHA I/II) were as follows: TAVR, 84.6%; SAVR, 63.3%; balloon valvuloplasty, 18.2%; medication, 21.4% (TAVR vs. SAVR, P = 0.04; SAVR vs. balloon valvuloplasty or medical therapy, P = 0.01). On multivariate regression analysis, renal failure (hazard ratio [HR] = 5.3, P < 0.001), not performing TAVR (HR = 4.9, P < 0.001), and pulmonary pressure (10 mm Hg, HR = 1.2, P = 0.02) were independent predictors of 1-year mortality. CONCLUSIONS TAVR, performed in carefully selected high-risk patients, is associated with an excellent survival rate and high functional class. Patients treated with another of the available modalities, including SAVR, had a worse outcome, regardless of which alternative treatment they receive.