Globally, breast cancer (BCa) incidence is rising and is driven largely by an increase in estrogen receptor (ER)+ cancers. The breast epithelium consists of three populations, including basal cells, as well as two subtypes of luminal cells – ER+ hormone sensing and secretory cells. Though it is known that these populations are maintained by lineage restricted progenitors, the molecular profile of these progenitor populations is yet to be described in humans. We set out to determine if the proposed cell-of-origin for ER+ breast cancer, ER+ luminal progenitors (LPs), were affected by breast cancer risk factors such as parity and hormone exposure. ER+ LPs were increased in nulliparous women as well as in control mice and those exposed to estrogen supplementation. They were decreased in oophorectomized or anti-estrogen treated mice. We describe the transcriptional signature of ER+ LPs in mouse and used this to annotate the analogous population in the human breast. Through this analysis we demonstrate that these cells are enriched for chromatin modifying enzymes, PI3K signaling, and genes that are commonly mutated in ER+ BCa. We also demonstrate that these molecular profiles are mirrored in ER+ tumours in multiple independent dataset. ER+ LPs are present in relatively high numbers in PIK3CA mutated ER+ BCa in women and in mouse models of ER+ BCa. A novel lineage tracing model demonstrated that PI3K activation in ER+ mammary tissue leads to dedifferentiation to a luminal/basal intermediate cell type prior to expansion of luminal progenitors. These results provide a mechanistic insight into ER+, Pik3ca-driven BCa, paving the way for research into developing precision ER+ BCa preventatives.