SUMMARY: This Danish cross-sectional study (n=20,905) showed that women aged 65-81 years generally underestimated fracture risk compared to absolute risk estimated by the FRAX® algorithm. Significant association was found between risk factors (e.g., previous fracture, parental hip fracture, and self-rated heath) and self-perceived fracture risk. Although women recognized the importance of some fracture risk factors, a number of significant risk factors appeared to be less well known. INTRODUCTION: The aim of this study is to investigate women's self-perceived fracture risk and potential factors associated with this and to compare self-perceived risk with absolute fracture risk estimated by FRAX® in women aged 65-80 years. METHODS: Data from 20,905 questionnaires from the ROSE study were analyzed. The questionnaire included 25 items on osteoporosis, risk factors for fractures, and self-perceived risk of fractures and enabled calculation of absolute fracture risk by FRAX®. Data were analyzed using bivariate tests and regression models. RESULTS: Women generally underestimated their fracture risk compared to absolute risk estimated by FRAX®. Women with risk factors for facture estimated their fracture risk significantly higher than their peers. No correlation between self-perceived risk and absolute risk was found. The ordered logistic regression model showed a significant association between high self-perceived fracture risk and previous fragility fracture, parental hip fracture, falls, self-rated heath, conditions related to secondary osteoporosis, and inability to do housework. CONCLUSIONS: These women aged 65-81 years underestimated their risk of fracture. However, they did seem to have an understanding of the importance of some risk factors such as previous fractures, parental hip fracture and falls. Risk communication is a key element in fracture prevention and should have greater focus on less well-known risk factors. Furthermore, it is important to acknowledge that risk perception is not based solely on potential risk factors but is also affected by experiences from everyday life to personal history.