The rates of fracture at sites with different relative amounts of cortical and trabecular bone (hip, spine, distal radius) have been used to make inferences about the pathomechanics of bone loss and the existence of type I and type II osteoporosis. However, fracture risk is directly related to the ratio of tissue stress to tissue strength, which in turn is dependent not only on tissue composition but also tissue geometry and the direction and magnitude of loading. These three elements determine how the load is distributed within the tissue. As a result, assumptions on the relative structural importance of cortical and trabecular bone, and how these tissues are affected by bone loss, can be inaccurate if based on regional tissue composition and bone density alone. To investigate the structural significance of cortical and trabecular bone in the proximal femur, and how it is affected by bone loss, we determined the stress distributions in a normal and osteoporotic femur resulting from loadings representing: (1) gait; and (2) a fall to the side with impact onto the greater trochanter. A three-dimensional finite element model was generated based on a representative femur selected from a large database of femoral geometries. Stresses were analyzed throughout the femoral neck and intertrochanteric regions. We found that the percentage of total load supported by cortical and trabecular bone was approximately constant for all load cases but differed depending on location. Cortical bone carried 30% of the load at the subcapital region, 50% at the mid-neck, 96% at the base of the neck and 80% at the intertrochanteric region. These values differ from the widely held assumption that cortical bone carries 75% of the load in the femoral neck and 50% of the load at the intertrochanteric region. During gait, the principal stresses were concentrated within the primary compressive system of trabeculae and in the cortical bone of the intertrochanteric region. In contrast, during a fall, the trabecular stresses were concentrated within the primary tensile system of trabeculae with a peak magnitude 4.3 times that present during gait. While the distribution of stress for the osteoporotic femur was similar to the normal, the magnitude of peak stress was increased by between 33% and 45%. These data call into question several assumptions which serve as the basis for theories on the pathomechanics of osteoporosis. In addition, we expect that the insight provided by this analysis will result in the improved development and interpretation of non-invasive techniques for the quantification of in vivo hip fracture risk.