Abstract Background The reverse left ventricular remodeling (LV-R) is well known phenomenon that may occurs in a wide spectrum of heart disease, generally associated with an improvement of prognosis. Recently, conflicting data are reported from two randomized trials assessing prognosis after MitraClip implantation. Purpose To assess time, magnitude and pattern of LV-R and its relationship with outcome, and to assess the role of ejection fraction (EF) in selection of patients for MitraClip procedure. Methods and results Among 47 patients treated with MitraClip for severe mitral regurgitation (MR), 2 switched to surgical intervention. All patients underwent echo before, at discharge, 1 and 6-month after procedure (age 78.2±8.3 years, NYHA 3.74±0.44, LV EF 36.5±12.8%, logistic EuroSCORE I 22.41±8.4, STS-PROM 4.6±1.9, functional MR 82%). From baseline to 6-month reverse LV-R was defined as a ≥15% decrease in LV end-systolic volume (LVESV) and an adverse LV-R as a increase of ≥10% in LVESV, respectively. At 6-month, sustained reduction of MR ≤2 was observed in all patients, but 2; reverse LV-R occurred in 51% (23), adverse LV-R in 18% (8) and no LV-R in 31% (14) patients. In a multivariate regression model, baseline LVEDV were a strong independent predictor of reverse LV remodeling [β −0.564, 95% CI: −0.363 to −0.074; P=0.004], whereas baseline LVEF was not (P=0.126). Furthermore, an LVEDV ≤130 mL was strongly associated with reverse remodeling with an OR: 0.796 (CI: 0.052–0.792, P=0.022). During follow-up (17.5±9.3 months), in adverse/no LV-R patients mortality for any cause and hospitalization for heart failure (HF) occurred in 50% vs. 95.7% (log-rank, P value= 0.006). By Cox analysis, adverse LV-R was strongly associated with mortality for any cause and hospitalization for HF with adjusted OR of 5.6 (95% CI: 1.65–19.00, P=0.006). Finally, combining adverse/no LV-R together the risk of mortality for any cause and hospitalization for HF increased with adjusted OR of 10.08 (95% CI: 1.29–78.6, P=0.027). Conclusion The half of real-world patients undergoing percutaneous mitral valve repair for severe MR showed reverse remodeling. However, there was a half of patients in whom afterload mismatch resulted in early and sustained adverse and no remodeling associated with subsequently high mortality and recurrence of HF. Baseline larger LV volumes rather than EF may help us to refine selection patients for MitraClip procedure, avoiding futility. Funding Acknowledgement Type of funding source: None