Although androgen is considered an anabolic hormone, the consequences of androgen receptor (AR) overexpression in skeletally-targeted AR-transgenic lines highlight the detrimental effect of enhanced androgen sensitivity on cortical bone quality. A compartment-specific anabolic response is observed only in male but not female AR3.6-transgenic (tg) mice, with increased periosteal bone formation and calvarial thickening. To identify anabolic signaling cascades that have the potential to increase bone formation, qPCR array analysis was employed to define expression differences between AR3.6-tg and wild-type (WT) periosteal tissue. Notably, categories that were significantly different between the two genotypes included axonal guidance, CNS development and negative regulation of Wnt signaling with a node centered on stem cell pathways. Further, fine mapping of AR3.6-tg calvaria revealed that anabolic thickening in vivo is not uniform across the calvaria, occurring only in frontal but not parietal bones. Multipotent fraction 1 progenitor populations from both genotypes were cultured separately as frontal bone neural crest stem-like cells (fNCSC) and parietal bone mesenchymal stem-like cells (pMSC). Both osteoblastic and adipogenic differentiation in these progenitor populations was influenced by embryonic lineage and by genotype. Adipogenesis was enhanced in WT fNCSC compared to pMSC, but transgenic cultures showed strong suppression of lipid accumulation only in fNCSC cells. Osteoblastogenesis was significantly increased in transgenic fNCSC cultures compared to WT, with elevated alkaline phosphatase (ALP) activity and induction of mineralization and nodule formation assessed by alizarin red and von Kossa staining. Osteocalcin (OC) and ALP mRNA levels were also increased in fNCSC cultures from AR3.6-tg vs. WT, but in pMSC cultures ALP mRNA levels, mineralization and nodule formation were decreased in AR3.6-tg cells. Expression differences identified by array in long bone periosteal tissue from AR3.6-tg vs. WT were recapitulated in the fNCSC samples while pMSCs profiles reflected cortical expression. These observations reveal the opposing effects of androgen signaling on lineage commitment and osteoblast differentiation that is enhanced in cells derived from a neural crest origin but inhibited in cells derived from a mesodermal origin, consistent with in vivo compartment-specific responses to androgen. Combined, these results highlight the complex action of androgen in the body that is dependent on the embryonic lineage and developmental origin of the cell. Further, these data these data suggest that the periosteum surrounding long bone is derived from neural crest.