The purpose of the study is to determine the effects of the BIG 1-98 treatments on bone mineral density. BIG 1-98 compared 5-year adjuvant hormone therapy in postmenopausal women allocated to four groups: tamoxifen (T); letrozole (L); 2-years T, 3-years L (TL); and 2-years L, 3-years T (LT). Bone mineral density T-score was measured prospectively annually by dual energy X-ray absorption in 424 patients enrolled in a sub-study after 3 (n = 150), 4 (n = 200), and 5 years (n = 74) from randomization, and 1 year after treatment cessation. Prevalence of osteoporosis and the association of C-telopeptide, osteocalcin, and bone alkaline phosphatase with T-scores were assessed. At 3 years, T had the highest and TL the lowest T-score. All arms except for LT showed a decline up to 5 years, with TL exhibiting the greatest. At 5 years, there were significant differences on lumbar T-score only between T and TL, whereas for femur T-score, differences were significant for T versus L or TL, and L versus LT. The 5-year prevalence of spine and femur osteoporosis was the highest on TL (14.5 %, 7.1 %) then L (4.3 %, 5.1 %), LT (4.2 %, 1.4 %) and T (4 %, 0). C-telopeptide and osteocalcin were significantly associated with T-scores. While adjuvant L increases bone mineral density loss compared with T, the sequence LT has an acceptable bone safety profile. C-telopeptide and osteocalcin are useful markers of bone density that may be used to monitor bone health during treatment. The sequence LT may be a valid treatment option in patients with low and intermediate risk of recurrence.