Background: No randomized study powered to compare balloon-expandable (BE) with self-expanding (SE) transcatheter heart valve (THV) on individual endpoints after transcatheter aortic valve replacement (TAVR) has been conducted to date. Methods: From January 2013 to December 2015, the FRANCE-TAVI nationwide registry included 12,141 patients undergoing BE-THV (Edwards, n=8038) or SE-THV (Medtronic, n=4103) for native aortic stenosis (AS). Long-term mortality status was available in all patients (median 20 months, IQR:14-30). Patients treated with BE-THV (n=3910) were successfully matched 1:1 with 3910 patients treated with SE-THV by using propensity-score (25 clinical, anatomical and procedural variables) and by date of the procedure (within 3 months). The first co-primary outcome was the occurrence of paravalvular regurgitation (PVR)≥moderate and/or in-hospital mortality. The 2 nd co-primary outcome was 2-year all-cause mortality. Results: In matched-propensity analyses, the incidence of the 1st co-primary outcome was higher with SE-THV (19.8%) compared with BE-THV(11.9%; RR=1.68; 95%CI:1.46-1.91; p<0.0001). Each component of the outcome was also higher in SE-THV patients: PVR≥moderate (15.5% vs. 8.3%; RR=1.90; 95% CI:1.63-2.22; p<0.0001) and in-hospital mortality (5.6% vs 4.2%, RR=1.34; 95%CI:1.07-1.66; p=0.01). During follow-up, all-cause mortality occurred in 899 patients treated with SE-THV (2-year mortality was 29.8%) and in 801 patients treated with BE-THV (2-year mortality 26.6%; HR=1.17; 95% CI:1.06-1.29; p=0.003). Similar results were found using inverse probability of treatment weighting using propensity score analysis. Conclusions: The present study suggests that use of SE-THV was associated with a higher risk of PVR and higher in-hospital and 2-year mortality as compared with BE-THV. These data strongly support the need for a randomized trial sufficiently powered to compare head-to-head the latest generation of SE and BE-THV.