Despite surgery and systemic treatment, many breast cancer patients relapse due to metastasis. One reason for such poor clinical outcomes concerns current approaches to understanding breast cancer biology. Current paradigms are disproportionately focused on intrinsic genetic and epigenetic changes underlying tumorigenesis. However, it has become clear that tumor cell interactions with the host facilitate malignant progression. Extensive studies reveal that a dominant cell-extrinsic element of the breast tumor microenvironment is the macrophage, termed tumor-associated macrophage (TAM). Although high TAM densities have been associated with a poorer prognosis, this does not indicate that TAMs are always pro-tumorigenic. In fact, proportions of patients with high TAM densities exhibit significant longevity compared to those with little TAM infiltration. This functional dichotomy introduces the notion that TAMs can be reclassified into M1 (tumor-suppressing) or M2 (tumor-promoting) subtypes reminiscent of the CD4+ Th1-Th2 paradigm. It is noteworthy that current identification of human TAMs is based on the expression of a single phenotypic marker, CD68, which is inadequate to distinguish functional diversity. Thus, we will test the novel hypothesis that TAM phenotype reflects the balance of M1 to M2 subtypes, which is determined by expression of IRF8, a key myeloid-dependent transcription factor. IRF8, unlike other IRF members is involved in diverse stages of myelopoiesis and is indispensable for cytokines/chemokines (e.g., IL-12, IL-18, CCL5) essential for antitumor immunity. Additional rationale for exploring IRF8 status in TAM biology is based on our recent work in myeloid-derived suppressor cells (MDSC) biology, a newly identified tumor-promoting myeloid population. Here, we showed that IRF8 expression significantly inhibited MDSC expansion, consistent with the mode of action of IRF8 in myelopoiesis. To test our central hypothesis, we propose three aims: 1) to determine the causal link between IRF8 expression and TAM phenotype in mouse mammary tumor models; 2) to determine how polarizing cytokine signals impede or repress IRF8 expression to impact TAM phenotype; and 3) to determine whether stratification of human TAMs into IRF8hi and IRF8lo-expressing subtypes improves prognostic significance in breast cancer. Altogether, we posit that separation of TAMs into subtypes reflecting distinct IRF8-reactivities will not only offer a mechanistic basis for their functional diversity, but also illuminate new ways to track patient outcomes.