Background— The Transcatheter Valve Therapy (TVT) registry model was recently developed to predict the risk of in-hospital mortality in patients undergoing transcatheter aortic valve replacement. We sought to externally validate the model in an independent data set of consecutively enrolled patients in the Swiss Transcatheter Aortic Valve Implantation registry. Methods and Results— The original prediction model was retrospectively applied to 3491 consecutive patients undergoing transcatheter aortic valve replacement in Switzerland between February 2011 and February 2016. We examined model performance in terms of discrimination (Harrel C index) and calibration (Hosmer–Lemeshow goodness-of-fit test) for prediction of in-hospital and 30-day mortality and compared its predictive accuracy with the Society of Thoracic Surgeons Predicted Risk of Mortality score. Rates of in-hospital and 30-day mortality in the external validation cohort were 2.9% and 3.8%, respectively. The TVT registry model was found to have moderate discrimination (C index, 0.66; 95% confidence interval, 0.60–0.72 and C index, 0.67; 95% confidence interval, 0.62–0.72 for in-hospital and 30-day mortality, respectively) and good calibration. Compared with the Society of Thoracic Surgeons Predicted Risk of Mortality score, the TVT registry model demonstrated improved calibration for in-hospital (slope, 0.83; P =0.23 versus slope, 0.24; P <0.001, respectively) and 30-day (slope, 1.11; P =0.40 versus slope, 0.41; P <0.001, respectively) mortality. Conclusions— In a large, multicenter, non-US cohort of patients with transcatheter aortic valve replacement, the validation of the TVT registry model demonstrated moderate discrimination and good calibration for the prediction of in-hospital and 30-day mortality. As a result, the TVT registry model should be considered an alternative to the Society of Thoracic Surgeons Predicted Risk of Mortality score for decision making and assessment of early outcome in patients eligible for transcatheter aortic valve replacement.