Acute myeloid leukemia (AML) initiating and sustaining cells maintain high cell-surface similarity with their cells-of-origin, i.e., hematopoietic stem and progenitor cells (HSPCs), and identification of truly distinguishing leukemia-private antigens has remained elusive to date. To nonetheless utilize surface antigen-directed immunotherapy in AML, we here propose targeting both, healthy and malignant human HSPC, by chimeric antigen receptor (CAR) T-cells with specificity against CD117, the cognate receptor for stem cell factor. This approach should spare most mature hematopoietic cells and would require CAR T termination followed by subsequent transplantation of healthy HSPCs to rescue hematopoiesis. We successfully generated anti-CD117 CAR T-cells from healthy donors and AML patients. Anti-CD117 CAR T-cells efficiently targeted healthy and leukemic CD117-positive cells in vitro. In mice xenografted with healthy human hematopoiesis, they eliminated CD117-expressing, but not CD117-negative human cells. Importantly, in mice xenografted with primary human CD117-positive AML, they eradicated disease in a therapeutic setting. Administration of ATG in combination with rituximab, which binds to the co-expressed CAR T-cell transduction/selection marker RQR8, led to CAR T-cell depletion. Thus, we here provide the first proof of concept for the generation and preclinical efficacy of CAR T-cells directed against CD117-expressing human hematopoietic cells.