Background: The impact of a change in mitral regurgitation (MR) following TAVR is unknown. We studied the impact of baseline MR and early post-procedural change in MR on survival following TAVR. Methods: The SWEDEHEART registry included all TAVRs performed in Sweden. Patients were dichotomized into no/mild and moderate/severe MR groups. Vital status, echocardiographic data at baseline and within 7 days after TAVR were analyzed. Results: 1712 patients were included. 1404 (82%) had no/mild MR and 308 (18%) had moderate/severe MR. Baseline moderate/severe MR conferred a higher mortality rate at 5-year follow-up (adjusted HR 1.29, CI 1.01-1.65, p = 0.04). Using persistent amp;lt;= mild MR as the reference, when moderate/severe MR persisted or if MR worsened from amp;lt;= mild at baseline to moderate/severe after TAVR, higher 5-year mortality rates were seen (adjusted HR 1.66, CI 1.17-2.34, p = 0.04; adjusted HR 1.97, CI 1.29-3.00, p = 0.002, respectively). If baseline moderate/severe MR improved to = mild after TAVR no excess mortality was seen (HR 1.09, CI 0.75-1.58, p = 0.67). Paravalvular aortic regurgitation (PVL) was inversely associated with MR improvement after TAVR (OR 0.4, 95%: CI 0.17-0.94; p = 0.034). Atrial fibrillation (OR 2.1, 95% CI: 1.27-3.39, p = 0.004), self-expanding valve (OR 3.8, 95% CI: 2.08-7.14, p amp;lt; 0.0001), and PVL (4.3, 95% CI 2.32-7.78. p amp;lt; 0.0001) were associated with MR worsening. Conclusions: Moderate/severe baseline MR in patients undergoing TAVR is associated with a mortality increase during 5 years of follow-up. This risk is offset if MR improves to amp;lt;= mild, whereas worsening of MR after TAVR is associated with a 2-fold mortality increase. (C) 2019 Elsevier B.V. All rights reserved.