Tumor progression and dissemination critically depend on support from the tumor microenvironment, the ensemble of cellular and acellular components surrounding and interacting with tumor cells. The extracellular matrix (ECM), the complex scaffolding of hundreds of proteins organizing cells in tissues, is a major component of the tumor microenvironment. It orchestrates cellular processes including proliferation, migration, and invasion, that are highly dysregulated during cancer progression. Alterations in ECM abundance, integrity, and mechanical properties have been correlated with poorer prognosis for cancer patients. Yet the ECM proteome, or "matrisome," of tumors remained until recently largely unexplored. This review will present the recent developments in computational and proteomic technologies that have allowed the comprehensive characterization of the ECM of different tumor types and microenvironmental niches. These approaches have resulted in the definition of protein signatures distinguishing tumors from normal tissues, tumors of different stages, primary from secondary tumors, and tumors from other diseased states such as fibrosis. Moreover, recent studies have demonstrated that the levels of expression of certain genes encoding ECM and ECM-associated proteins is prognostic of cancer patient survival and can thus serve as biomarkers. Last, proteomic studies have permitted the identification of novel ECM proteins playing functional roles in cancer progression. Such proteins have the potential to be exploited as therapeutic targets.