OBJECTIVES: Cost-effectiveness analyses are currently conducted after the conduct of expensive clinical trials. An approach was developed to estimate cost-effectiveness thresholds prior to clinical studies, comparable to statistical power calculations. METHODS: A new osteoporosis treatment was taken as example, with different scenarios of treatment efficacy and costs. Data on fracture and mortality risks were obtained from the General Practice Research Database. These risks were estimated individually by age, sex, fracture history, body mass index, smoking and other risk factors. EQ5D utilities were obtained from a UK national report (NICE) and outcomes were simulated over a 10-year period (5-year treatment), using a cost-acceptability ratio of £30k per QALY gained. RESULTS: The 5-year risk of osteoporotic fracture required to reach the cost-effectiveness threshold was 17.1% (95% confidence interval 15.0-19.3%) with a fracture efficacy of 0.50 at an annual cost of £1000. This was 6.1% (5.2-7.0%) with a cost of £250 and 3.7% (3.1-4.5% with a cost of £100. With a fracture efficacy of only 0.80, these threshold risks were 34.7% (23.2-38.7%), 10.7% (8.3-14.5%) and 5.4% (3.8-7.8%), respectively. At a T-score of -2.5 and fracture efficacy of 0.80 and cost of £250, patients without a fracture history would require additional risk factors (with a relative rate of 2.5) in order to reach the threshold, while this would be reached by the average woman at age 85. However, with a cost of £1000, this threshold would only be reached at age 55 with additional risk factors with RR of 9.5 and at age 85 with RR of 3.0. CONCLUSION: Cost-effectiveness thresholds can be estimated prior to expensive clinical trials using high-quality healthcare databases. Similar to statistical power calculations, they can then be used to guide patient selection into the clinical trials, by providing information on the required minimum levels of risks.