Objective:To analyze and evaluate the different methods in detecting anti-mitochondrial antibodies (AMA) for the diagnostic performance of primary biliary cholangitis (PBC).Methods:In this retrospective study, we collected serum samples of 37 PBC patients who were admitted to the Second Hospital of Jiaxing, from November 2022 to November 2023, 43 patients with other autoimmune diseases and 53 normal controls during the same time. The results of AMA were detected in parallel by indirect immunofluorescence assay (IIF), immunoblot tests (IB) and chemilluminescent microparticle immunoassay (CLIA). Fisher′s exact test was used to compare the count data between different groups, while a paired chi-square test was used to assess the results of the three test methods. We analyzed the performance of the three test methods; Venn diagram was used to show the consistency distribution of AMA results, the diagnostic performance of AMA for PBC were evaluated by area under receiver operating characteristic curve (ROC).Results:The positive rate of AMA in PBC group was significantly higher than that in the other two groups, and AMA was detected by CLIA which had higher positive rate (86.5%, 32/37) than the other two methods. As shown in the Venn diagram, the proportion of consistent AMA results among the three methods was 74.4%(99/133), and between IB and CLIA methods which used for detecting AMA-M2, the consistent proportion was 86.5%(115/133). Among comparison of the performance of three methods, the highest specificity (92.7%)was IIF method, and the highest sensitivity (86.5%) was CLIA method. The respective AUC of IIF, IB and CLIA methods were 0.801, 0.800 and 0.875. In the combined detection, the AUC of IIF+CLIA method was 0.942, which was better than the other two or triple combined method detection.Conclusion:For highly suspected PBC patients with negative IIF screening test, combined IB or CLIA could be used to detect AMA-M2. We recommended that clinicians should refer to the AMA results conducted by two or three methods, and preferentially recommended the IIF method combined with CLIA method for the detection of AMA.