Bone density measurements are frequently used to access the quality of a person's bone. However, bone density values do not give complete information on the bone's mechanical characteristics which ultimately determine the likelihood of bone fracture. PURPOSE: To investigate the relationship between bone density, bone strength and trabecular bone architecture. METHODS: Twelve femurs were harvested from six mature rabbits. Each femur's bone mineral density (BMD) was measured using DXA. Femur compression strength was determined using an Instron with a sample of the distal end of one femur per rabbit. Two thin cross-sections of the other femur were prepared for imaging with a scanning electron microscope of three regions (medial, middle and lateral) per sample. BoneJ software was used to compute various measurements of trabecular architecture for each region: ratio of bone volume to trabecular volume, trabecular thickness, and trabecular separation. For each of these three values the median value between the three regions was used for analysis. Pearson correlation coefficients were analyzed (p < 0.05). RESULTS: Strength and stiffness measurements were positively correlated (r = 0.995), while bone ratio and trabecular separation were negatively correlated (r = -0.849). BMD and trabecular thickness were positively correlated (r = 0.792), but BMD and strength were not. CONCLUSIONS: High bone mineral density values do not necessarily indicate good bone quality. Additionally, it was confirmed that trabecular architecture is not uniform within a cross-section or even within a specific region of that cross-section. This further raises the concern for using BMD values to make conclusions about a particular bone's quality when it is clear that trabecular bone is not a homogenous tissue. Supported by NSF grant # NSF-UBM 0634592.