Background Autoantibodies are useful biomarkers for the early detection and diagnosis of autoimmune cerebellar ataxia (ACA).Objective To identify novel autoantibody candidates in ACA patients.Methods Patients with cerebellar ataxia of unknown cause were recruited from July 2018 to February 2023. Anti-neural autoantibodies in patient samples were detected by tissue-based indirect immunofluorescence assay (TBA) on rat cerebellum sections. TBA-positive samples were further screened for well-established anti-neural autoantibodies using commercial kits. Tissue-immunoprecipitation (TIP) and subsequent mass spectrometric (MS) analysis were used to explore the target antigens of autoantibodies in samples that were TBA-positive but negative for known autoantibodies. The specific binding between autoantibodies and the identified target antigen was confirmed by neutralization experiments, recombinant cell-based indirect immunofluorescence assay (CBA), and western blotting experiments.Results The eukaryotic translation elongation factor 1 delta (EEF1D) protein was identified as a target antigen of autoantibodies in samples from a 43-year-old female ACA patient, while the specific binding of autoantibodies and EEF1D was confirmed by subsequent experiments. A second anti-EEF1D autoantibody-positive ACA patient, a 59-year-old female, was detected in simultaneous screening. The main clinical manifestations in each of the two patients were cerebellar syndrome, such as unsteady walking and limb ataxia. Both patients received immunotherapy, including corticosteroids, intravenous immunoglobulin, and mycophenolate mofetil. Their outcomes provided evidence to support the effectiveness of immunotherapy, but the cerebellar atrophy that occurred before treatment may be irreversible.Conclusion In the current study, we identified anti-EEF1D autoantibody as a novel autoantibody candidate in ACA. Its pathological roles and diagnostic value need to be further verified in larger-scale studies.