The mechanical competence of bone depends upon its quantity, structural arrangement, and chemical composition. Assessment of these factors is important for the evaluation of bone integrity, particularly as the skeleton remodels according to external (e.g. mechanical loading) and internal (e.g. hormonal changes) stimuli. Micro magnetic resonance imaging (µMRI) has emerged as a non-invasive and non-ionizing method well-suited for the repeated measurements necessary for monitoring changes in bone integrity. However, in vivo image-based directional dependence of trabecular bone (TB) has not been linked to mechanical competence or fracture risk despite the existence of convincing ex vivo evidence. The objective of this dissertation research was to develop a means of capturing the directional dependence of TB by assessing a fabric tensor on the basis of in vivo µMRI. To accomplish this objective, a novel approach for calculating the TB fabric tensor based on the spatial autocorrelation function was developed and evaluated in the presence of common limitations to in vivo µMRI. Comparisons were made to the standard technique of mean-interceptlength (MIL). Relative to MIL, ACF was identified as computationally faster by over an order of magnitude and more robust within the range of the resolutions and SNRs achievable in vivo. The potential for improved sensitivity afforded by isotropic resolution was also investigated in an improved µMR imaging protocol at 3T. Measures of reproducibility and reliability indicate the potential of images with isotropic resolution to provide enhanced sensitivity to orientation-dependent measures of TB, however overall reproducibility suffered from the sacrifice in SNR. Finally, the image-derived TB fabric tensor was validated through its relationship with TB mechanical competence in specimen and in vivo µMR images. The inclusion of trabecular bone fabric measures significantly improved the bone volume fraction-based prediction of elastic constants calculated by micro-finite element analysis. This research established a method for detecting TB fabric tensor in vivo and identified the directional dependence of TB as an important determinant of TB mechanical competence.