R ecently, the indication for transcatheter aortic valve replacement (TAVR) has been expanded not only to patients with degenerative aortic stenosis considered inoperable or of high surgical risk but also to those with intermediate surgical risk. [1][2][3] Thus, optimal patient risk stratification should be performed before the TAVR procedure. Frailty, which is not captured in the classical surgical risk model, is considered highly prevalent in elderly vulnerable patients and can be characterized by Background-Gait speed reflects an important factor of frailty and is associated with an increased risk of late mortality in patients with cardiac disease. This study sought to assess the prognostic value of gait speed in elderly patients who underwent transcatheter aortic valve replacement. Methods and Results-We investigated the 5-m or 15-feet gait speed (m/sec) in 1256 patients who underwent transcatheter aortic valve implantation using data from the OCEAN-TAVI Japanese multicenter registry (Optimized Catheter Valvular Intervention-Transcatheter Aortic Valve Implantation). Baseline characteristics, procedural outcomes, and all-cause mortality were compared among groups defined by differential gait speed classification: model 1, normal (>0.83 m/sec; n=563; 44.8%), slow (0.5-0.83 m/sec; n=429; 34.2%), slowest (<0.83 m/sec; n=205; 16.3%), unable to walk (n=48; 3.8%); and model 2, classification and regression tree survival model indicating the threshold of gait speed as 0.385 m/sec (>0.385 m/sec; n=1080 versus ≤0.385 m/sec; n=117). The cumulative 1-year mortality rate showed significant differences in the classical gait speed groups in model 1 (7.6%, 6.6%, 18.2%, and 40.7%, respectively; P<0.001) and survival classification and regression tree group in model 2 (7.7% versus 21.9%; P<0.001). The slowest walkers and those unable to walk demonstrated independent associations with increased midterm mortality after adjustment for several confounding factors (hazard ratio, 1.83, 4.28; 95% confidence interval, 1.03-3.26, 2.22-8.72; P=0.039, <0.001, respectively). Gait speed <0.385 m/sec determined by classification and regression tree also independently associated with worse prognosis (hazard ratio, 2.40; 95% confidence interval, 1.75-5.88; P=0.001). Conclusions-Gait speed using both traditional and specific classification is useful as a potential marker for predicting vulnerable patients associated with adverse clinical outcomes after transcatheter aortic valve replacement. (Circ Cardiovasc Interv. 2017;10:e005088.