OBJECTIVES: The purpose of this study was to estimate the cost-effectiveness of alendronate therapy for osteopenic postmenopausal women in Japan. METHODS: A Markov model with six health states (no fracture, post-vertebral fracture, post-hip fracture, post-vertebral and hip fracture, bedridden, and death) was developed to predict lifetime costs and quality-adjusted life years (QALYs) of five years of alendronate therapy versus no drug treatment in postmenopausal women without fracture history. Fracture risk associated with age and bone mineral density (BMD) was derived from epidemiologic studies in Japan. We ran the model with different combinations of age (65 to 75), BMD (70% - 80% of the young adult mean (YAM)), and the number of clinical risk factors (CRFs, one to three). Probabilistic sensitivity analysis was performed to assess parameter uncertainty. RESULTS: The model was sensitive to age, BMD, and CRFs. The incremental cost-effectiveness ratio (ICER) was below $50,000 per QALY in the following scenarios: 1) In 70-year-old women with BMD 70% of YAM, who had two CRFs, 2) In 75-year-old women with BMD 70% of YAM, who had two CRFs. In 65-year-old women with BMD 70% of YAM, who had three CRFs, 3) In 70-year-old women with BMD 70% and 75% of YAM, who had three CRFs, and 4) In 75-year-old women with BMD 70% and 75% of YAM, who had three CRFs. Applying a willingness to pay threshold of $50,000 per QALY, the probability of being cost-effective was estimated to 2.9 %, 36.5%, and 99.2% in 70-year-old women with BMD 70% of YAM with one CRF, two CRFs, and three CRFs, respectively. CONCLUSIONS: Whether to treat osteopenic postmenopausal Japanese women should be determined with consideration for multiple CRFs in terms of cost-effectiveness.