Children with metastatic osteosarcoma have an extremely poor prognosis, but little is known about the biologic variables that contribute to metastasis formation. Tumor cells can be found in the circulation of patients with disease burden, but which cells migrate into circulation, and their relationship to the primary tumor and metastatic disease is unknown. This question can not be addressed by traditional bulk sequencing methods that mask the heterogeneity of malignant cells within tumors. Furthermore, it is known that failure of immune surveillance facilitates metastasis, but whether the tumor microenvironment (TME) in osteosarcoma correlates to metastatic disease is unknown. We aim to build a cellular atlas of pediatric osteosarcoma to 1) describe the tumor and immune microenvironment of osteosarcoma using single cell technologies, 2) analyze individual circulating tumor cells (CTCs) by single cell sequencing to describe their phenotypic and phylogenetic relationship to primary and metastatic disease and 3) to correlate tumor and immune single cell gene expression patterns with clinical behavior and biologic features such as genomic alterations. We hypothesize that the TME of osteosarcoma is correlated with clinical variables such as presence of metastasis and response to therapy. By studying circulating tumor cells in the context of primary or metastatic tumors, we aim to demonstrate relationship of CTCs to a heterogeneous tumor, and thereby define the subpopulation with metastatic potential. We will create a cell by cell atlas of primary and metastatic osteosarcoma and CTCs, overlaid with clinical and biologic features. From this atlas we will discover biomarkers predictive of clinical behavior. Knowledge of the TME and features of cells with metastatic potential will inform clinical trials of novel therapeutics that could translate into improved cures for children with osteosarcoma.