This study aims at exploring biventricular remodelling and its implications for outcome in a representative patient cohort with severe aortic stenosis (AS) undergoing transcatheter aortic valve implantation (TAVI). Pre-interventional echocardiographic examinations of 100 patients with severe AS undergoing TAVI were assessed by speckle tracking echocardiography of both ventricles. Association with mortality was determined for right ventricular global longitudinal strain (RVGLS), RV free wall strain (RVFWS) and left ventricular global longitudinal strain (LVGLS). During a median follow-up of 1,367 [959-2,123] days, 33 patients (33%) died. RVGLS was lower in non-survivors [-13.9% (-16.4 to -12.9)] than survivors [-17.1% (-20.2 to -15.2); P = 0.001]. In contrast, LVGLS as well as the conventional parameters LV ejection fraction (LVEF) and RV fractional area change (RVFAC) did not differ (P = ns). Kaplan-Meier analyses indicated a reduced survival probability when RVGLS was below the -14.6% cutpoint (P < 0.001). Lower RVGLS was associated with higher mortality [HR 1.13 (95% CI 1.04-1.23); P = 0.003] independent of LVGLS, LVEF, RVFAC, and EuroSCORE II. Addition of RVGLS clearly improved the fitness of bivariable and multivariable models including LVGLS, LVEF, RVFAC, and EuroSCORE II with potential incremental value for mortality prediction. In contrast, LVGLS, LVEF, and RVFAC were not associated with mortality. In patients with severe AS undergoing TAVI, RVGLS but not LVGLS was reduced in non-survivors compared to survivors, differentiated non-survivors from survivors, was independently associated with mortality, and exhibited potential incremental value for outcome prediction. RVGLS appears to be more suitable than LVGLS for risk stratification in AS and timely valve replacement.