Osteoporosis is a major public health problem. For the time being, the diagnosis of osteoporosis relies on densitometry (T-score < -2.5 by DXA), although the risk of fracture depends also on other factors than the bone mass. Osteoporosis diagnosis (DXA) must be distinguished from the individual risk assessment of fracture. Different risk factors complementary to bone mass have been already validated in different populations. These include an old age, a history of fracture after the age of 50, a familial history of hip fracture (father or mother), a low BMI (< 20), corticoid treatment (> 3 months), tabagism and excessive alcohol consumption. A WHO taskforce has combined these different factors in order to integrate them in a 10-years predictive risk model of fracture (FRAX**). This model should still be validated in different populations, especially in populations not included in its development, which is the case for Belgium. We are evaluating these different risk factors for fracture in a Brussels population of 5000 women (60-80 years) who will be followed each year during 10 years. We also assess the predictive value of other risk factors for fracture not included in the WHO model (tendency to fall, use of sleeping pills, early non substituted menopause, sedentarity, ...). In an interim analysis of the first 452 women included and with data yet available at the time of this writing, we could find a significant (P < 0.05) relationship between diagnosis of osteoporosis at DXA and the number of risk factors, age > 70 years, a personal history of fracture after 50 years and a BMI < 20.