Despite the success of the aromatase inhibitors (AIs) in treating estrogen receptor positive breast cancer, 15-20 % of patients receiving adjuvant AIs will relapse within 5-10 years of treatment initiation. Long-term estrogen deprivation (LTED) of breast cancer cells in culture mimics AI-induced estrogen depletion to dissect mechanisms of AI resistance. However, we hypothesized that a subset of patients receiving AI therapy may maintain low circulating concentrations of estrogens that influence the development of endocrine resistance. We expanded established LTED models to account for incomplete suppression of estrogen synthesis during AI therapy. MCF-7 cells were grown in medium with charcoal-stripped serum supplemented with defined concentrations of 17β-estradiol (E2) or the estrogenic androgen metabolite 5α-androstane-3β,17β-diol (3βAdiol), an endogenous selective estrogen receptor modulator. Cells were selected in concentrations of E2 or 3βAdiol that induce 10 or 90 percent of maximal proliferation (EC(10) and EC(90), respectively), or estrogen deprived. Estrogen independence was evaluated during selection by assessing cell growth in the absence or presence of E2 or 3βAdiol. Following >7 months of selection, estrogen independence developed in estrogen-deprived cells and EC(10)-selected cells. Functional analyses demonstrated that estrogen-deprived and EC(10)-selected cells developed estrogen independence via unique mechanisms, ERα-independent and dependent, respectively. Estrogen-independent proliferation in EC(10)-selected cells could be blocked by kinase inhibitors. However, these cells were resistant to kinase inhibition in the presence of low steroid concentrations. These data demonstrate that further understanding of the total estrogen environment in patients on AI therapy who experience recurrence is necessary to effectively treat endocrine-resistant disease.