Neonatal hypoxic-ischemic encephalopathy is a major cause of brain damage in infants, and is associated with periventricular white matter injury and chronic neurological dysfunctions. However, the mechanisms of the chronic white matter injury and reorganization are still unclear. In this study, in vivo diffusion tensor imaging (DTI) was employed to evaluate the late changes of white matter microstructural integrity in the rat brains at 10 weeks after severe neonatal hypoxic-ischemic insults at postnatal day 7. In the fractional anisotropy directionality map, qualitative evaluation showed that a dorsoventrally oriented fiber bundle extended from the corpus callosum into the cyst in the anterior brain, whilst the posterior peri-infarct areas had similar fiber orientations as the contralateral internal capsule, optic tract and fimbria of hippocampus. Compared to the contralateral hemisphere, significantly higher fractional anisotropy, axial diffusivity and diffusion trace value were observed quantitatively in the distal end of the extended fiber bundle connecting the anterior and posterior white matters rostrocaudally. A significantly lower fractional anisotropy but higher axial and radial diffusivities and trace were also found in the ipsilateral corpus callosum, proximal external capsule and anterior commissure, while slightly lower fractional anisotropy and axial diffusivity were noticed in the ipsilateral internal capsule and optic nerve. It was suggested that increased fractional anisotropy, axial diffusivity and trace characterize white matter reorganization in chronic neonatal hypoxic-ischemic insults, whereas reduction in fractional anisotropy appears to characterize two types of white matter lesions, with significantly higher axial and radial diffusivities and trace being primary and slightly lower axial diffusivity being secondary. Combined with fractional anisotropy directionality map, in vivo DTI provides important indices to differentiate the chronic effects of severe neonatal hypoxic-ischemic injury and recovery globally, quantitatively and non-invasively.