Abstract Background/Aims Ankylosing spondylitis (AS) is a chronic inflammatory arthritis leading to long term disability. It is frequently diagnosed late when irreversible damage has already occurred. Unlike most autoimmune rheumatological diseases, there are no known autoantibodies associated with AS; this can confound the diagnostic challenge as diagnosis depends not only on symptoms but on finding evidence of active inflammation at the time of assessment or damage from previously active disease. Our aim was to determine the extent to which autoantibodies may be pathognomonic of AS. Methods A systematic literature review was conducted to identify all articles reporting on autoantibodies in AS. The protocol was preregistered on PROSPERO. Articles were screened and reviewed independently by two reviewers. Disagreements were solved by consensus. Data were extracted and the MINORS (Methodological Index for Non-Randomised Studies) tool was used to assess and compare the quality of the studies. Data were pooled in a narrative synthesis. Results 743 papers were identified after removal of duplicates. Following review of abstracts, 31 full text articles were assessed for eligibility and 18 were excluded, leaving 13 studies to be included in the final systematic review. 12 articles looked at 14 biomarkers; one article looked at 2 high density nucleic acid protein arrays expressing 3,498 proteins. Samples sizes were small and the papers were found to be of modest quality. AS patients showed a wide-ranging autoantibody response across studies. One study found that 60% of autoantibodies detected were found only in the AS cohort versus the rheumatoid arthritis cohort and heathy controls. Antibodies against HLA-B27, pANCA, CD74, OmpC, collagens, PPM1A, noggin, sclerostin, Klebsiella pneumoniae and Breg cells were found more commonly in AS patients than in controls (both healthy controls and those with other autoimmune diseases). Antibodies to human tTG and anti DFS70 were not found to be associated with AS and there was conflicting evidence regarding the association of antibodies to the cell-wall mannan of Saccharomyces cerevisiae (ASCA) in AS from different papers. Conclusion Although many autoantibodies have been found among persons with AS, there is currently no evidence that any are specifically and independently associated with the disease. We therefore conclude that as yet no autoantibodies can be considered to be pathognomonic for AS. However, some autoantibodies may be clinically significant and combined with other biomarkers, such as HLA-B27, may prove fruitful in reducing the delay to diagnosis. Further work should seek to explore other avenues, in particular autoantibodies in closely related diseases, such as inflammatory bowel disease, to try to determine autoantibodies that may aid in the earlier diagnosis of this chronic disease. Disclosure E.J. Davies: None. G.T. Jones: None. R. Sengupta: Honoraria; Received grants, honoraria, and expenses for attendance at advisory board meetings or conferences and for giving lectures from AbbVie, Biogen, Celgene, Lilly, MSD, Novartis, Roche and UCB. Grants/research support; Received grants, honoraria, and expenses for attendance at advisory board meetings or conferences and for giving lectures from AbbVie, Biogen, Celgene, Lilly, MSD, Novartis, Roche and UCB.