The purpose of this study was to evaluate the feasibility and initial results of a multidisciplinary percutaneous aortic valve implantation (PAVI) program including the transfemoral approach (TFA) and the transapical approach (TAA). This was a prospective registry including all patients with symptomatic severe aortic stenosis who were evaluated for PAVI and those who finally underwent the procedure. All patients were considered very high risk or nonsurgical candidates, and an algorithm based on prespecified criteria was used to determine the most appropriate approach (TFA vs TAA) for each patient. The Edwards-Sapien valve was used in all cases, and all procedures were performed without cardiopulmonary bypass. A total of 29 consecutive patients were evaluated for PAVI, and 22 of them underwent the procedure (mean age 84 +/- 7 years, predicted surgical mortality 26 +/- 16%), with 11 patients selected for each approach. Reasons for selecting TAA rather than TFA were the following: small diameter and/or severe calcification of the iliofemoral arteries (4 patients), peripheral vascular disease (4 patients), porcelain aorta (2 patients), and horizontal ascending aorta (1 patient). Successful aortic valve implantation was obtained in 91% of the procedures, and procedural and 30-day mortality were 4.3% and 8.7%, respectively. There were no cases of myocardial infarction, vascular complications, or cerebrovascular accident. PAVI was associated with a significant reduction in mean aortic gradient (baseline 34 +/- 10 mm Hg, postprocedure 9 +/- 3 mm Hg, p <0.001), with no cases of severe aortic regurgitation. After a median follow-up period of 6 months, there were no additional major adverse cardiac events, and no significant changes in hemodynamic aortic valve parameters were observed. In conclusion, this study demonstrates the feasibility and low complication rate of a PAVI program using a prospective strategy for the selection of the TFA or TAA, with half of the patients selected for each approach.