Objective The impact of early allograft dysfunction on the outcome after liver transplantation has yet to be established. We explored the independent predictive value of the Model for End-Stage Liver Disease (MELD) score measured in the post-transplant period on the risk of mortality or retransplantation. Material and methods Retrospective cohort study on adults undergoing orthotopic deceased donor liver transplantation from 2004 to 2014. The MELD score was determined prior to transplantation and daily until 21 days after. The risk of mortality or re-transplantation within the first year was assessed according to quartiles of MELD using unadjusted and adjusted stepwise Cox regression analysis. Results We included 374 consecutive liver transplant recipients of whom 60 patients died or were re-transplanted. The pre-transplant MELD score was comparable between patients with good and poor outcome, but from day 1 the MELD score significantly diversified and was higher in the poor outcome group (MELD score quartile 4 versus quartile 1-3 at day 10: HR 5.1, 95% CI: 2.8-9.0). This association remained after adjustment for non-identical blood type, autoimmune liver disease, and hepatocellular carcinoma (adjusted HR 5.3, 95% CI: 2.9-9.5 for MELD scores at day 10).The post-transplant MELD score was not associated with pre-transplant MELD score or the Eurotransplant donor risk index. Conclusion Early determination of the MELD score as an indicator of early allograft dysfunction after liver transplantation was a strong independent predictor of mortality or re-transplantation and was not influenced by the quality of the donor, or preoperative recipient risk factors.