Uveitis, a group of intraocular inflammatory diseases, is one of the major causes of severe visual impairment among the working-age population. This study aimed to screen potential urinary biomarkers for uveitis based on proteome analysis. An experimental autoimmune uveitis (EAU) rat model induced by bovine interphotoreceptor retinoid-binding protein (IRBP) was used to mimic uveitis. In discovery phase, a total of 704 urinary proteins were identified via data-independent acquisition (DIA) proteomic technique, of which 76 were significantly changed (34, 36, and 37 on days 5, 8, and 12, respectively, after bovine IRBP immunization). Gene Ontology annotation of the differential proteins indicates that acute-phase response, innate immune response, neutrophil aggregation, and chronic inflammatory response were significantly enriched. Protein-protein interaction network indicates that these differential urinary proteins were biologically connected in EAU, as a group. In validation phase, 17 proteins having human orthologs were verified as the potential markers associated with uveitis by parallel reaction monitoring (PRM) targeted quantitative analysis. Twelve differential proteins changed even when there were no clinical manifestations or histopathological ocular damage. These 12 proteins are potential biomarkers for early diagnosis of uveitis to prevent the development of visual impairment. Five differential proteins changed at three time-points and showed progressive changes as the uveitis progressed, and another five differential proteins changed only on day 12 when EAU severity peaked. These 10 proteins may serve as potential biomarkers for prognostic evaluation of uveitis. Our findings revealed that the urinary proteome could sensitively reflect dynamic pathophysiological changes in EAU, and represent the first step towards the application of urinary protein biomarkers for uveitis.