OBJECTIVES: Evidence-based guidelines recommend treatment for postmenopausal women with osteopo­rosis to prevent fractures. The study aims at determining since when this was known and whether the utilization of bisphosphonates in Germany from this point onward was adequate and to what extent health gains might have been foregone due to a limited use of bisphos­phonates. METHODS: To determine since when the beneficial effect (prevention of fractures) was known, cumu­lative meta-analyses of randomized controlled trials derived from systematic reviews were conducted. The evidence-base was considered as established, when a significant (5%-level) reduction of fractures was observed in trial populations combined in meta-analysis compared to therapies without bisphosphonates. Utilization figures for bisphosphonates and epidemio­logical estimates where taken from published sources. RESULTS: The hip/femur fracture risk was significantly lower if treatment included bisphosphonates compared to treatment without bisphosphonates (RR 0.62; 95%-CI 0.40-0.97 / RR 0.45; 95%-CI 0.23-0.90). In principal, this was known since 1995/1996. Utilization of bisphospho­nates in 1996 was sufficient for the continuous treatment of about 8,200 patients (440,000 patients in 2006). About 1.6 to 1.9 million patients annually might have benefitted from treatment. About 22,800 fractures might thus have been avoided, had all patients with potential benefit continuously received bisphosphonates since 1996/1997. CONCLUSIONS: The delay in the wider use of bisphosphonates for osteoporosis treatment has resulted in a considerable loss of potential health gains in terms of avoided fractures. An arguable lack of evidence for the expected benefit from bisphosphonate therapy does not sufficiently explain this finding. Other factors (e. g. cost considerations) might have contributed to this result. Limitations of the present analysis are primarily associated with uncertainties of epidemiolo­gical estimates and the application of study results to the entire patient population.