Tauopathies consist of over 25 different neurodegenerative diseases that include argyrophilic grain disease (AGD), progressive supranuclear palsy (PSP), corticobasal degeneration (CBD), and Pick’s disease (PiD). Tauopathies are defined by brain accumulation of microtubule-associated protein tau in fibrillar aggregates, whose prevalence strongly correlates with dementia. Dominant mutations in tau cause neurodegenerative diseases, and most increase its aggregation propensity. Pathogenesis of tauopathies may involve pathological tau conformers that serve as templates to recruit native protein into growing assemblies and also move between brain cells to cause disease progression, similar to prions. Prions adopt pathological conformations, termed “strains,” that stably propagate in living systems, and create unique patterns of neuropathology. Data from multiple laboratories now suggest that tau acts as a prion. It propagates unique strains indefinitely in cultured cells, and when these are inoculated into mouse models, they create defined neuropathological patterns, which establish a direct link between conformation and disease. In humans, distinct fibril structures are associated with different diseases, but causality has not been established as in mice. Cryo-EM structures of tau fibrils isolated from tauopathy brains reveal distinct fibril cores across disease. Interestingly, the conformation of the tau monomer unit within different fibril subtypes from the same patient appears relatively preserved. This is consistent with data that the tau monomer samples an ensemble of conformations that act as distinct pathologic templates in the formation of restricted numbers of strains. The propensity of a tau monomer to adopt distinct conformations appears to be linked to defined local motifs that expose different patterns of amyloidogenic amino acid sequences. The prion hypothesis, which predicts that protein structure dictates resultant disease, has proved particularly useful to understand the diversity of human tauopathies. The challenge now is to develop methods to rapidly classify patients according to the structure of the underlying pathological protein assemblies to achieve more accurate diagnosis and effective therapy.