Historically, defining haematopoietic subsets, including self-renewal, differentiation and lineage restriction, has been elucidated by transplanting a small number of candidate cells with many supporting bone marrow (BM) cells. While this approach has been invaluable in characterising numerous distinct subsets in haematopoiesis, this approach is arguably flawed. The haematopoietic stem cell (HSC) has been proposed as the critical haematopoietic subset necessary for transplantation. However, due to the presence of supporting cells, the HSC has never demonstrated sufficiency. Utilising the homeobox B5 (Hoxb5)-reporter system, we found that neither long-term (LT) HSCs nor short-term (ST) HSCs alone were sufficient for long-term haematopoietic reconstitution. Critically, reconstitution can be rescued by transplanting combined LT-and ST-HSCs, without supporting cells; a fraction we term the 'Minimum Subset for Transplantation' (MST). The MST accounts for only 0Á005% of nucleated cells within mouse BM, and this MST can be cultured, expanded and genetically modified while preserving its rapid haematopoietic engraftment potential. These results support the consideration of an MST approach for clinical translation, especially for gene therapy approaches that require HSC compartment modification.