AIM: To evaluate the predictive value of islet autoantibodies for the diagnosis of autoimmune uveitis (AU), as well as to characterize the association bet ween islet autoantibodies and AU. METHODS: Totally 97 patients with AU and 100 healthy persons without any autoimmune diseases as the control group were recruited. Multiple serum islet autoantibodies were measured using commercial enzyme-linked immunosorbent assay kits (ELISA). A supplementary questionnaire was used to complement the subject’s demographics and clinical features. The level of glucose concentrations and white blood cells were measured. Conditional logistic regression was performed to estimate odds ratios (ORs), and 95% confidence intervals (CIs) of AU according to islet autoantibodies and to evaluate the predictive value of islet autoantibodies for AU diagnosis. Autoantibodies subgroups and other variables were included into analysis. RESULTS: In AU patients, the prevalence of detecting at least one of the autoantibodies was 31.9% (31/97). The most frequent autoantibody was ZnT8A (30.9%), followed by GADA (11.3%), IA-2A (4.1%), ICA (2.1%) and IAA (2.1%). Islet autoantibodies were found to be correlated positively with AU diagnosis [OR (95%CI): 13.86 (3.28, 58.50), P<0.001]. Moreover, Zn-T8A was remarkably correlated with AU diagnosis [OR (95%CI): 6.13 (1.96, 19.17), P<0.001], In contrast, neither GADA nor other islet antibodies (IA-2A, ICA and IAA) showed any association with AU risk under an additive model. CONCLUSION: The prevalence of islet antibodies, especially ZnT8A, in patients with AU is higher. Islet antibodies as well as novel biomarkers should be included in routine evaluation at AU and is a valuable biological marker to classify newly-diagnosed uveitis.