[Truncated abstract] Areal bone mineral density (BMD) is measured routinely in the clinic by a quantitative radiological technique, dual-energy X-ray absorptiometry (DXA). BMD is used widely to assess non-invasively but indirectly the mechanical fragility of bone and consequently is able to predict fracture risk. While BMD correlates well with in vitro measurements of bone strength it does not directly measure a mechanical property; half of incident minimally traumatic fractures in women occur with BMD values above the World Health Organisation defined threshold for osteoporosis. This arises partly because the mechanical strength of bone is dependent on its structural geometry and material strength as well as bone mineral mass. Essentially, bones fracture when load stresses exceed the mechanical capacity of the material to withstand them. The structural geometry (i.e., the amount of bone tissue and its complex three-dimensional arrangement within the macroscopic bone envelope) defines the stresses produced by a given load, while the intrinsic load capacity of the material is defined by the composition and microstructure of the bone tissue itself. Hip structural analysis (HSA) is a technique that elucidates the structural geometric component of bone strength; essentially combining information available from conventional DXA images of the proximal femur with a biomechanical beam model based on the stresses arising in a combination of pure bending and axial compression. A version of HSA has recently been released commercially, and has obtained US Food and Drug Administration approval for its clinical application. ... Given the acknowledged limitations of the HSA method when applied to 2-D projection images, a 3-D approach to structural geometry, using imaging modalities such as pQCT and QCT or a recently introduced version of DXA that mimics QCT, is indicated for the future. With that in mind and the possibility of the anthropometric phantom being adopted for future accuracy and precision assessments, improvements in the design of this phantom are recommended. Studies to better understand and verify Contents v the relevance of the 'local buckling' phenomenon as a structural geometric factor in the genesis of macro-fractures are also recommended. In summary, it is essential that superior (compared to BMD) non-invasively determined clinical predictors of bone fragility leading to fracture be investigated. Structural geometric variables are potential candidates. This has led to consideration of; (i) the need to progress beyond BMD for a more sensitive and specific bone strength measurement; (ii) theoretical advantages of structural geometry over BMD; (iii) limitations of the current HSA technique based on DXA, including those introduced by its restrictive assumptions; (iv) the value of HSA in longitudinal studies, exemplified by the 'normal' but rapid skeletal changes seen in human lactation, with possible implications for an analogous study of the menopause; and (v) an investigation, using a custom-designed anthropometric phantom, of the adaptation of HSA to certain emerging imaging modalities and methods able to resolve bone structural geometry in three dimensions.