Transcatheter aortic valve replacement (TAVR) is an established therapy improving outcome in patients with severe aortic stenosis (AS). Cardiovascular magnetic resonance (CMR)-derived left atrioventricular coupling index (LACI) has demonstrated prognostic value, however, its prognostic utility in severe AS remains unknown. Between January 2017 and September 2023 138 consecutive patients with severe AS (80 years (70-83), 62% male) referred for TAVR were prospectively recruited for pre-procedural CMR imaging. LACI was defined as the ratio of the left atrial (LA) end-diastolic volume index (LA EDVi) and left ventricular (LV) end-diastolic volume index (LV EDVi). The primary endpoint was cardiovascular (CV)-mortality. The cohort was dichotomized at a median of 44.1% (high vs. low LACI). Patients with increased LACI (≥44.1%) had higher symptom burden (NYHA III/IV prevalence (68.1% vs. 44.9%; P = 0.047)), more frequently atrial fibrillation (50.7% vs. 14.5%; P < 0.001), and elevated NT-proBNP (2017 ng/L vs. 1012 ng/L; P = 0.007). Over a median follow-up of 2.7 years (IQR 1.7-3.6), high LACI was associated with higher CV mortality (log-rank P = 0.016). In exploratory multivariable Cox regression models, LACI remained associated with CV-mortality after adjustment for established parameters of left atrial or ventricular function (LA reservoir strain: HR 2.19, 95% CI 1.05-4.57, P = 0.036; LV GLS: HR 2.21 95% CI 1.00-4.9, P = 0.049). In patients with severe aortic stenosis, CMR-derived LACI was associated with cardiovascular mortality and may serve as a structural marker of advanced atrioventricular remodelling. Given its simplicity and routine availability in standard CMR workflows, LACI may serve as a clinically practical risk marker for baseline stratification.