Myelodysplastic syndromes (MDS) are clonal stem cell disorders marked by bone marrow failure with varying degrees of cytopenias. MDS are characterized by the accumulation of complex genetic alterations that influence disease pathogenesis and outcomes. Given the heterogeneity of MDS patients, multiple prognostic scoring systems have been developed to predict patient outcomes. Most focus primarily on untreated patients in whom outcomes are predicted accurately only at diagnosis; their application dynamically, during the disease course and especially after hypomethylating agent (HMA) failure, may not be accurate. In this focused review, we discuss the utility of established MDS prognostic models and a novel tool at the time of HMA failure, along with the strengths and limitations of each.