Study DesignRetrospective study.PurposeTo assess the feasibility and limitations of fracture risk assessment tool (FRAX) for osteoporotic vertebral fractures in the Korean population.Overview of LiteratureThe FRAX algorithm is country specific and uses clinical risk factor data to calculate an individual patient's 10-year probability of hip fracture and 10-year probability of major osteoporotic fracture. However, it has not been adequately investigated for Korean.MethodsOne hundred ninety four patients who had all risk factor data for the calculation of FRAX were divided into two groups depending on the existence of vertebral fractures: the fracture group was comprised of 88 patients and the non-facture group comprised of 105 patients. We analyzed prediction of the fracture by applying respectively the Korean, Japanese, USA and UK model, and compared their FRAX results by calculating lumbar bone mineral density (BMD) instead of femoral neck BMD.ResultsThe prediction of vertebral fracture using FRAX was 10.9 ± 6.2% in the fracture group, 9.5 ± 5.5% of the non-fracture group in the Korean model (p = 0.108); 17.9 ± 10.2% in the fracture group, 14.6 ± 9.0% in the non-fracture group in the Japanese model (p = 0.017). Only the Japanese model exhibited significant difference in vertebral fracture risk. The prediction of vertebral fracture using lumbar BMD instead of femoral neck BMD was 19.5 ± 12.1% in the fracture group, 16.0 ± 10.3% in the non-fracture group in the Korean model (p = 0.029). All models had statistically significant differences for the prediction of osteoporotic vertebral fracture.ConclusionsThe 10-year probability of osteoporotic vertebral fracture had underestimation of the risk considering treatment eligibility based on the National Osteoporosis Foundation guidelines. BMD that accurately reflects the contribution of each result to fracture risk should be preferred for the prediction of fracture using FRAX, when lumbar spine and hip BMD measurements are both performed for clinical purposes in Korean.