Diffusion weighted magnetic resonance imaging (DW-MRI), provides unique insight into the microstructure of neural white matter tissue, allowing researchers to more fully investigate white matter disorders. The abundance of clinical research projects incorporating DW-MRI into their acquisition protocols speaks to the value this information lends to the study of neurological disease. However, the most widespread DW-MRI technique, diffusion tensor imaging (DTI), possesses serious limitations which restrict its utility in regions of complex white matter. Fueled by advances in DW-MRI acquisition protocols and technologies, a group of exciting new DW-MRI models, developed to address these concerns, are now becoming available to clinical researchers. The emergence of these new imaging techniques, categorized as high angular resolution diffusion imaging (HARDI), has generated the need for sophisticated computational neuroanatomic techniques able to account for the high dimensionality and structure of HARDI data. The goal of this thesis is the development of such techniques utilizing prominent HARDI data models. Specifically, methodologies for spatial normalization, population atlas building and structural connectivity have been developed and validated. These methods form the core of a comprehensive analysis paradigm allowing the investigation of local white matter microarcitecture, as well as, systemic properties of neuronal connectivity. The application of this framework to the study of schizophrenia and the autism spectrum disorders demonstrate its sensitivity sublte differences in white matter organization, as well as, its applicability to large population DW-MRI studies.