The magnitude of osteoporosis, the established relationship between lo
w bone mass and the risk of fracture, and the availability of preventi
ve treatment suggest that the early detection of women with low bone m
ass is justified. The feasibility of population screening using bone m
ass measurements remains controversial. Another approach is the use of
clinical risk factors to detcct women at high risk. However, several
studies have demonstrated that the assessment of risk factor status do
es not appear to be an efficient tool for the identification of perime
nopausal women with low bone mass. The poor performance of the predict
ion models might be explained in part by unmeasured factors, especiall
y genetic factors, which are an important determinant of bone mass. On
the other hand, the clinical usefulness of clinical risk factors need
s to be more precisely evaluated, especially in the detection of women
at high risk for hip fracture.