Progressive supranuclear palsy (PSP) is a common neurodegenerative disorder causing progressive motor, oculomotor, bulbar, and cognitive deficits. Current treatments do not arrest disease progression and the prognosis for functional independence and survival is correspondingly poor. Pathologically, PSP is characterized by accumulation of insoluble hyperphosphorylated forms of the 4‐repeat isoform of the microtubule associated protein tau (4R‐Tau) and subsequent loss of CNS neurons. Convergent lines of evidence suggest that 4R‐Tau is mechanistically involved in the neurodegeneration of PSP. Clinical and pathological PSP phenocopies can result from splice site mutations in the MAPT gene causing over‐expression of 4R‐Tau. Furthermore, PSP is strongly associated with the H1 MAPT haplotype that gives rise to relative over‐production of 4R‐Tau. Consequently, compounds that prevent neurodegeneration caused by 4R‐Tau in vivo may be effective therapeutic agents in PSP.The long‐term goal of our work is to develop effective treatments for progressive supranuclear palsy (PSP). Development of therapies for PSP and other tauopathies has been limited by lack of animal models that allow discovery‐based approaches, for example screening candidate chemical modifiers against disease‐relevant phenotypes in vivo. The zebrafish is a genetically and chemically tractable vertebrate model that is particularly suited to drug discovery and comparative efficacy testing and is being used increasingly for studying human neurological diseases. The objective of this study was to develop a zebrafish model of PSP and to carry out a phenotype based chemical modifier screen in vivo.We generated transgenic zebrafish that express human 4R‐Tau in CNS neurons. Tau transgenic zebrafish replicated many clinic‐pathological features of PSP, including: hypokinesia; oculomotor deficits; somato‐dendritic accumulation of hyperphosphorylated tau; neuronal apoptosis; and impaired survival. Importantly, these phenotypes were quantitative, replicable and occured rapidly allowing their use as endpoints in a phenotype‐driven screen. We tested over 200 bioactive compounds to identify chemicals that rescued motor function in the 4R‐Tau zebrafish, using a custom automated video tracking and motion analysis method. We identified several ‘hits’ that rescued motor function and showed activity in subsequent verification assays.Our completion of the first unbiased phenotype‐based chemical modifier screen in a vertebrate tauopathy model provides proof‐of‐concept that drug discovery can be carried out in vivo by assaying disease‐relevant neurological deficits. Importantly our screen has so far identified several candidates for further development as possible tau therapeutics.Support or Funding InformationUS Department of Veterans' Affairs (BX003168‐01)NINDS (NS080881)CurePSP (441‐05, 468‐08)Pittsburgh Foundation (AD201577642)