Abstract Myelofibrosis (MF) belongs to the family of classic Philadelphia-negative myeloproliferative neoplasms (MPNs). It can be primary myelofibrosis (PMF) or secondary myelofibrosis (SMF) evolving from polycythemia vera (PV) or essential thrombocythemia (ET). Despite the differences, PMF and SMF patients are currently managed in the same way, and prediction of survival is based on the same clinical and genetic features. In the last few years, interest has grown concerning the ability of gene expression profiles (GEPs) to provide valuable prognostic information. Here, we studied the GEPs of granulocytes from 114 patients with MF, using a microarray platform to identify correlations with patient characteristics and outcomes. Cox regression analysis led to the identification of 201 survival-related transcripts characterizing patients who are at high risk for death. High-risk patients identified by this gene signature displayed an inferior overall survival and leukemia-free survival, together with clinical and molecular detrimental features included in contemporary prognostic models, such as the presence of high molecular risk mutations. The high-risk group was enriched in post-PV and post-ET MF and JAK2V617F homozygous patients, whereas pre-PMF was more frequent in the low-risk group. These results demonstrate that GEPs in MF patients correlate with their molecular and clinical features, particularly their survival, and represent the proof of concept that GEPs might provide complementary prognostic information to be applied in clinical decision making.