The clinical inefficiency of cancer immunotherapy is in part due to the presence of an immunosuppressive network that favors tumor progression; in fact, by producing soluble molecules such as cytokines, interleukins and growth factors, tumors induce an alternative hematopoiesis that modifies the normal myeloid cell differentiation, pushing proliferation and expansion of cells with immunosuppressive function called myeloid-derived suppressor cells (MDSCs). The MDSC presence and frequency in blood of tumor patients is often reported as a prognostic marker that correlates with the clinical outcome and response to therapy. These cells use distinctive and redundant pathways to suppress the proliferation and function of antigen-stimulated T lymphocytes. Although MDSCs are heterogeneous, it appears that three main immunosuppressive cell subsets can be identified: granulocytic, monocytic and more immature cells, which might be able to originate the other two subpopulations. Our knowledge about mechanisms used by MDSC to restrain adaptive and innate immunity, the possibility to generate MDSCs by in vitro culture of bone marrow precursors, the definition of transcription factors and microRNAs regulating their in vivo expansion and maturation allow to hypothesize and design novel strategies for MDSC use as regulators of disorders characterized by excessive or uncontrolled stimulation of the immune response, such as transplant rejection and autoimmune diseases.