Background: Osteoporosis and fractures are long term complications of aromatase inhibitor use in the adjuvant therapy of breast cancer. Early detection of patients (pts) at risk of treatment-induced osteoporosis may allow preventive therapy and selection of the most appropriate endocrine therapy. We developed a statistical model describing the evolution of BMD in pts treated with T, L and sequences of T and L. Methods: Available dual-energy x-ray absorptiometry exams (DXA) of pts treated in trial BIG 1-98 were retrospectively collected from Swiss centers. Treatment arms: A) T for 5 years; B) L for 5 years; C) 2 years of T followed by 3 years of L; and D) 2 years of L followed by 3 years of T. Pts without DXA were used as a control for detecting selection bias. A repeated measures model using the first-order autoregressive covariance structure to allow for different times between DXA was used to model BMD (g/cm2) since trial randomisation. Prospectively defined covariates were considered as fixed effects in a multivariable model using an intention to treat analysis. Covariates at trial randomization were: age, height, weight, race, known osteoporosis, tobacco use, prior bone fracture, prior hormone replacement therapy (HRT), bisphosphonate use and previous neo-/adjuvant chemotherapy (ChT). Results: A total of 247 out of 546 pts had between 1 and 5 DXA; a total of 576 DXA were collected. Arm A contained 67 pts, B 63 pts, C 55 pts and D 62 pts. Median follow-up was 5.8 years. Factors correlated with BMD in the multivariate analysis were weight (0.003/kg, p < 0.0001), height (0.003/cm, p = 0.0083), osteoporosis (-0.130, p < 0.0001), tobacco (current / previously vs. never: -0.057, p = 0.0011 / -0.042, p = 0.0798), previous HRT (0.030, p = 0.0244), ChT (0.032, p = 0.0174), time since endocrine therapy start (-0.009/year, p = 0.0164) and treatment arm (B / C / D vs. A: -0.068, p = 0.0002 / -0.091, p < 0.0001 / -0.064, p = 0.003). Conclusions: Our statistical model describes the BMD evolution for pts treated with L and/or T. All treatment regimens affect BMD. Contrary to expectation, the switch schedule T followed by L does not seem to result in better bone protection compared to L monotherapy