The evolutionary origin of myelinating cells in the vertebrate nervous system remains a mystery. A clear delineation of the developmental potentialities of neuronal support cells in the CNS and PNS might aid in formulating a hypothesis about the origins of myelinating cells. Although a glial-precursor cell in the CNS can differentiate into oligodendrocytes (OLs), Schwann cells (SCs) and astrocytes, a homologous multipotential cell has not yet been found in the PNS. Here, we identify a cell type of embryonic dorsal root ganglia (DRG) of the PNS – the satellite cell – that develops into OLs, SCs and astrocytes. Interestingly, satellite-cell-derived OL precursors were found in cultures prepared from embryonic day 17 (E17) to postnatal day 8 (P8) ganglia, but not from adult DRGs, revealing a narrow developmental window for multipotentiality. We suggest that compromising the organization of the ganglia triggers a differentiation pathway in a subpopulation of satellite cells, inducing them to become myelinating cells with either a CNS or PNS phenotype. Our data provide an additional, novel piece in the myelinating-cell-precursor puzzle, and lead to the concept that cells in the CNS and PNS that function to ensheath neuronal cell bodies and axons can differentiate into OLs, SCs and astrocytes. In sum, it appears that glial fate might be determined over and above the CNS/PNS dichotomy. Last, we suggest that primordial ensheathing cells form the original cell population in which the myelination program first evolved.