Abstract Objective The objective of this study was to assess imaging predictors of mitral regurgitation (MR) improvement and to evaluate the impact of MR regression on long-term outcome in patients undergoing transcatheter aortic valve replacement (TAVR). Background Concomitant MR is a frequent finding in patients with severe aortic stenosis but usually left untreated at the time of TAVR. Methods Mitral regurgitation was graded by transthoracic echocardiography before and after TAVR in 677 consecutive patients with severe aortic stenosis. 2-year mortality was related to the degree of baseline and discharge MR. Morphological echo analysis was performed to determine predictors of MR improvement. Results 15.2% of patients presented with baseline MR ≥ 3 +, which was associated with a significantly decreased 2-year survival (57.7% vs. 74.4%, P < 0.001). MR improved in 50% of patients following TAVR, with 44% regressing to MR ≤ 2 +. MR improvement to ≤ 2 + was associated with significantly better survival compared to patients with persistent MR ≥ 3 +. Baseline parameters including non-severe baseline MR, the extent of mitral annular calcification and large annular dimension (≥ 32 mm) predicted the likelihood of an improvement to MR ≤ 2 +. A score based on these parameters selected groups with differing probability of MR ≤ 2 + post TAVR ranging from 10.5 to 94.4% (AUC 0.816; P < 0.001), and was predictive for 2-year mortality. Conclusion Unresolved severe MR is a critical determinant of long term mortality following TAVR. Persistence of severe MR following TAVR can be predicted using selected parameters derived from TTE-imaging. These data call for close follow up and additional mitral valve treatment in this subgroup. Graphic abstract Factors associated with MR persistence or regression after TAVR