Purpose Adjuvant therapy for breast cancer can be associated with decreased bone mineral density (BMD) that may lead to skeletal morbidity. This study examined whether zoledronic acid can prevent bone loss associated with adjuvant endocrine therapy in premenopausal patients. Patients and Methods This study is a randomized, open-label, phase III, four-arm trial comparing tamoxifen (20 mg/d orally) and goserelin (3.6 mg every 28 days subcutaneously) ± zoledronic acid (4 mg intravenously every 6 months) versus anastrozole (1 mg/d orally) and goserelin ± zoledronic acid for 3 years in premenopausal women with hormone-responsive breast cancer. In a BMD subprotocol at three trial centers, patients underwent serial BMD measurements at 0, 6, 12, 24, and 36 months. Results Four hundred one patients were included in the BMD subprotocol. Endocrine treatment without zoledronic acid led to significant (P < .001) overall bone loss after 3 years of treatment (BMD, −14.4% after 36 months; mean T score reduction, −1.4). Overall bone loss was significantly more severe in patients receiving anastrozole/goserelin (BMD, −17.3%; mean T score reduction, −2.6) compared with patients receiving tamoxifen/goserelin (BMD, −11.6%; mean T score reduction, −1.1). In contrast, BMD remained stable in zoledronic acid–treated patients (P < .0001 compared with endocrine therapy alone). No interactions with age or other risk factors were noted. Conclusion Endocrine therapy caused significant bone loss that increased with treatment duration in premenopausal women with breast cancer. Zoledronic acid 4 mg every 6 months effectively inhibited bone loss. Regular BMD measurements and initiation of concomitant bisphosphonate therapy on evidence of bone loss should be considered for patients undergoing endocrine therapy.