In the diagnosis of osteoporosis using single energy quantitative CT (SE-QCT) on the axial skeleton, only spongy bone mineral density (BMD) is used at present. Although the density of cortical bone is also determined by most QCT methods, it is not used for evaluation. The objective of this study was to determine the extent to which the cortical bone of the lumbar vertebral bodies accounts for their load-bearing capacity and failure behaviour, and to use this information to suggest improvements in the differential diagnosis of osteoporosis. Investigations were conducted in a clinical, theoretical-numerical and biomechanical-experimental context. Cortical (BMDC) and spongy (BMDS) bone mineral density was measured by SE-QCT/85 kV on 179 patients (68 males, 111 females). These bone densities were matched with the vertebral body fractures previously determined from conventional X-rays. A finite element model was used to study the variation in structural and material parameters of the vertebral body. 19 vertebral bodies that had been removed post-mortem were available for the biomechanical-experimental investigations. Spongy and cortical bone densities were also determined by SE-QCT on these vertebral bodies. Their failure load was then measured in the axial compression test. These investigations show that, in addition to the spongiosa, the cortical shell plays an important role in the load-bearing capacity of the vertebral body. If the spongiosa is weakened due to a loss of BMD, the residual load-bearing capacity of the vertebral bodies is increasingly shouldered by the cortical bone. The lower susceptibility to fracture in men compared with women when spongy bone mineral density is reduced can thus be attributed to the lack of a reduction in cortical bone mineral density. It is recommended that the BMDC also be evaluated in future, especially in the diagnosis of bone mass losses in women, to improve the estimation of the individual fracture risk.