Background Antisynthetase syndrome (ASSD) lacks of established clinico-serological classification criteria. A taskforce by EULAR and ACR is working on to develop and validate classification criteria for ASSD. Objectives To systematically upraise literature to retrieve the available definitions of ASSD and to evaluate their diagnostic performance. Methods This systematic review followed a pre-specified protocol. Two research questions (Q1: how is ASSD defined; Q2: what is the diagnostic performance of the definitions) were rephrased into PICOs terms to create search strategies. Studies on patients with suspect or confirmed ASSD, including a definition of the disease with any study design, excluding case-reports and narrative reviews, were eligible for inclusion. The diagnostic performance had to be tested against the reference standard of expert opinion. PubMed and Embase were searched from 01/01/1984 to 06/11/2018. Moreover, the ACR and EULAR congress abstracts (2017-2018) were hand searched. The titles and abstracts of the retrieved studies were screened by pairs of reviewers, the full-text of studies fulfilling the inclusion criteria was assessed to confirm eligibility. The references of the included studies were also evaluated in search of additional studies. Data from primary studies were extracted into a pre-specified extraction form and, if possible, 2x2 tables to assess diagnostic performance were completed. Sensitivities, specificities, positive and negative likelihood ratios (LR) were calculated for each study. If the diagnostic performance of a definition or variable was assessed in at least 4 studies, a meta-analysis of diagnostic performance was undertaken. The risk of bias (RoB) was assessed using the most appropriate tool depending on study design. Results After the exclusion of duplicates, the searches retrieved 4358 studies, of which 375 suitable for full-text review. Finally, 77 studies were included, along with 1 additional study from hand search and 3 congress abstracts. 72 studies were included in Q1 and 9 in both Q1 and 2. The presence of antisynthetase antibodies (70 studies), mainly anti-Jo1 (57 studies); myositis (51 studies), mainly defined clinically (32 studies); and interstitial lung disease (38 studies) were the variables most frequently used to define ASSD. Other variables, such as arthritis (19 studies), Raynaud’s phenomenon and skin manifestations (10 studies each) were included less frequently. Most commonly, ASSD was defined by a combination of clinical and serological variables. However, no study evaluated the diagnostic performance of such combined definitions. Most of the studies included in Q2 (6) evaluated specific variables of muscle biopsy, one evaluated MRI and 2 clinical variables, with a wide variability in the performance of each item. It was possible to meta-analyze data only to assess the performance of perifascicular necrosis/atrophy: pooled sensitivity (95%CI) was 0.53 (0.33,0.72) and specificity 0.63 (0.47,0.76), pooled LR+ 1.45 (0.72,2.89) and LR- 0.73 (0.40,1.34). Conclusion ASSD is defined according to a variety of combinations of serological, clinical and histological variables. The performance of these combined definitions however has not been tested, and from the limited evidence available single muscle biopsy variable (perifascicular necrosis/atrophy) seem to perform poorly. The systematic review confirms the need of data and consensus driven classification criteria for ASSD. Disclosure of Interests Giovanni Zanframundo: None declared, Sara Faghihi Kashani: None declared, Lorenzo Cavagna: None declared, Rohit Aggarwal: None declared, Carlomaurizio Montecucco Speakers bureau: AbbVie, Bristol-Myers Squibb, Celgene, Sanofi, Genzyme, Lilly, MSD, Pfizer, UCB, Carlo Alberto Scire: None declared, Garifallia Sakellariou: None declared