Uterine smooth muscle tumors of uncertain malignant potential (STUMPs) represent a heterogeneous group of tumors that cannot be histologically diagnosed as unequivocally benign or malignant. For this reason, many authors are working to obtain a better definition of diagnostic and prognostic criteria. In this work, we analyzed the genomic and epigenomic profile of uterine smooth muscle tumors (USMTs) in order to find similarities and differences between STUMPs, leiomyosarcomas (LMSs) and leiomyomas (LMs), and possibly identify prognostic factors in this group of tumors. Array-CGH data on 23 USMTs demonstrated the presence of a more similar genomic profile between STUMPs and LMSs. Some genes, such as PRKDC and PUM2, with a potential prognostic value, were never previously associated with STUMP. The methylation data appears to be very promising, especially with regards to the divergent profile found in the sample that relapsed, characterized by an overall CGI hypomethylation. Finally, the Gene Ontology analysis highlighted some cancer genes that could play a pivotal role in the unexpected aggressive behavior that can be found in some of these tumors. These genes could prove to be prognostic markers in the future.