Objective:We conducted a real-world multi-center clinical study with a large sample size to comprehensively evaluate the performance of three commercial hepatitis C virus (HCV) core antigen assays. The study aimed to evaluate the performance for their use in HCV infection screening, and to provide clues for further improving the sensitivity and specificity of the assays.Methods:Key performance indicators including the lower limit of detection (LOD), diagnostic sensitivity, and specificity of three HCV antigen assays (the Architect, Laibo, and ChemClin HCV core antigen assays) were evaluated using commercial seroconversion panels reflecting early HCV infection and clinical routine serum samples of outpatients and inpatients from 3 tertiary hospitals from January 2018 to April 2022. Factors that affect the performance indicators were further investigated.Results:The window period for detecting HCV infection with the three antigen assays was equal to or slightly longer than that of the RNA assay, but all are shorter than that of the anti-HCV assay. There was a good linear positive correlation between HCV core antigen and HCV RNA levels in treatment naive patients with hepatitis C ( r=0.90, P<0.01). For the most common genotype 1b strain in China, the LOD of the three HCV assays were equivalent to 531 IU/ml (Architect), 3,698 IU/mL (Laibo), and 4,624 IU/mL (ChemClin) HCV RNA, respectively. Due to the skewed distribution of HCV RNA levels in treatment-naive hepatitis C patients, more than 95% of the patients had viral loads higher than 6 166 IU/ml. Therefore, the three HCV antigens assays still maintained a satisfactory diagnostic sensitivity (94.33%-99.40%). Among 54 immunodeficient patients (leukemia patients) with HCV infection, 9% (5/54) had negative anti-HCV results, while the HCV antigen assays found all these infectors. Through further experiments, we revealed the amino acid polymorphism in the core region of genotype 3 strain impaired the sensitivity of all three HCV antigen assays. In addition, the sensitivity of the two domestic assays was impaired by anti-HCV antibodies in the serum. The specificity of HCV antigen assays for diagnosing hepatitis C is 99.94% to 99.98%. The rheumatoid factors, autoantibodies, and other unknown interference substances can lead to a small number of low level, "false positive" antigen results. Conclusions:HCV core antigen assay may be used as a satisfactory approach of infection screening, especially for the immunodeficient patents. However, the sensitivity and specificity of the assays are influenced by multiple factors, which should be further improved.