Neural tube and somites have long been thought to derive from separate germ layers: the ectoderm and mesoderm. This concept was challenged by the discovery of neuromesodermal progenitors, a bi-potent cell population that gives rise to both spinal neural tube and somites. In line with their proposed potency, these cells are considered to co-express the neural marker Sox2 and the mesodermal marker T/Brachyury. We performed genetic lineage tracing in mouse embryos and confirmed that T-expressing cells give rise to both neural tube and mesoderm. Surprisingly, however, Sox2-expressing cell derivatives colonise only the neural tube after embryonic day 8.5. Deletion of Sox2 in T-expressing cells was compatible with an otherwise normal neural tube and paraxial mesoderm. Moreover, Sox2 expression is absent from the chordoneural hinge, where neuro-mesodermal progenitors are located. Our findings demonstrate that neuro-mesodermal progenitors express T but notSox2, suggesting the need for re-evaluation of the neuro-mesodermal progenitor hypothesis.The initial aim of this study was to define the role of NMPs in neural tube formation, since it is not clear to what extent the progenitors contribute to this tissue. We traced cells either from the NSB or CLE into the closing neural tube, and found that the caudal end of the embryo harbours two distinct populations, both of which fulfil the criteria of NMPs, yet give rise to different domains within the closed neural tube. NMP function was addressed using laser ablation and a genetic approach, based on the proposed co-expression of T and Sox2.