OBJECTIVES: To assess the CE of risendronate vs. alendronate in post-menopausal women with established osteoporosis. METHODS: A fracture-incidence based Markov model was used to perform CEA from a health-care payer perspective, based on Polish data on health-care resource utilisation and unit cost. The basic transition probabilities have been taken for published epidemiological data, where possible from Poland. Efficacy of treatments in terms of ability to reduce the incidence of hip, vertebral, wrist fractures was populated from the systematic review and modelled by the use of relative risks in entering a transition state. Treatments were given for 5 years, assuming that the effect on fracture probability is instantaneous and persist unchanged through the treatment period. Separately women at the ages 50, 60, 70, or 80 years of age were analysed. Costs and clinical outcomes were observed from the time of therapy initiation until the end of treatment. The effectiveness frontier was mapped and the cost-effectiveness of each therapy was examined in terms of ICER. Extensive sensitivity analyses were undertaken. RESULTS: At each age-group risendronate was the most effective and most costly option, giving 17.6; 11.8; 6.5 and 3.08 QALY at 7707.7, 7745.7, 7421.5, and 6117 PLN (1 EURO=4 PLN) in 50, 60, 70, and 80 aged women, respectively. In the group at 50 and 70 alendronate was excluded because of extended dominance and the ICER for risendronate vs. no treatment was 137,651 and 71,084 PLN/QALY, respectively. The ICER for risendronate vs. alendronate in women at 60 and 80 was 267,393 and 46,732 PLN/QALY, respectively. These findings were fairly insensitive to variation in the main parameters.CONCLUSION: Considering the suggested threshold for cost-effectiveness, calculated on the basis of 1-year haemodialysis treatment cost (64,000 PLN) risendronate in comparison to alendronate appears to be cost-effective option in women at 70 and 80.