The model for end-stage liver disease (MELD) has been used as a predictor of mortality after left ventricular assist device (LVAD) placement. However, improvement or worsening of MELD and how those changes affect outcomes is unknown. We performed a retrospective analysis of 244 patients implanted with a continuous flow (CF) LVAD. Patients were dichotomized at admission into low- or high-risk categories using a cutoff of MELD ≥ 19, and they were reclassified at day of implant forming four groups: Group LL (low to low, remained low risk), LH (low to high, worsened to high risk), HH (high to high, remained high risk), and HL (high to low, improved to low risk). Patients who improved to a low risk (group HL) had the same 1 year survival as those that remained low risk (group LL; 80% vs. 77%; p = 0.6). However, patients who were initially classified as low risk and worsened to a high risk (group LH) had a survival that was worse than those that were consistently high risk (group HH; 55% vs. 10%; p = 0.01). Model for end-stage liver disease reclassification after adjusting for commonly attributed risk factors remained an independent predictor for mortality, including patients classified as Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) 1 and 2. In conclusion, our MELD score reclassification is an independent and powerful predictor of mortality in patients undergoing LVAD implantation.