Rabies diagnosis is mainly made on fresh brain tissue postmortem by means of the direct immunofluorescence test. However, in some cases, it is not possible to use this technique, given that the affected nervous tissue goes through a postmortem degradation process, due to problems in the handling and transport of the samples. For this reason, the preservation in time of the rabies virus inclusions was assessed, as well as the immunoreactivity and the ultrastructure of viral particles in tissue with postmortem degradation. Brains of mice inoculated with rabies virus and control mice were processed for conventional histology, immunohistochemistry, electron microscopy, and immunoelectron microscopy in different postmortem times. In the processed tissues for hematoxylin and eosin, the presence of eosinophilic inclusions was not observed beyond 12 h postmortem. Surprisingly, the immunoreactivity of the viral antigens increased with time, at least until 30 h postmortem. It was possible to easily recognize the viral particles by means of conventional electron microscopy until 12 h postmortem. Immunoelectron microscopy allowed us to identify the presence of viral antigens disseminated in the neuronal cytoplasm until 30 h postmortem, but immunoreactive viral particles were not observed. The rabies infection did not cause histological or ultrastructural alterations different from those in the control group as a result of the postmortem degradation. In conclusion, the immunohistochemistry is a reliable test for rabies diagnosis in samples with postmortem degradation and that have been fixed with aldehydes.