Background: In the early stage of COVID-19, the colloidal gold method was the first to be approved by the National Medical Products Administration for clinical use, but the interfering factors of false positive antibody test results are still unclear. Objective: To explore how to analyze and interpret the positive results of COVID-19 antibody test in clinical practice. Methods: A total of 8 678 patients simultaneously submitted for SARS-CoV-2 IgM/IgG antibody and nucleic acid in Peking University International Hospital from March to June 2020 for retrospective analysis were included in the inretrospective analysis. The epidemiological history, clinical manifestations and laboratory indicators(including SARS-CoV-2 IgM/IgG antibody and nucleic acid, rheumatoid factor, complement, immunoglobulin). The colloidal gold method was used to detect the SARS-CoV-2 IgM/IgG antibody, and the positive check of the antibody was performed by the magnetic particle chemiluminescence method. RT-PCR method was adopted to detect SARS-CoV-2 nucleic acid. The heterophilic antibody blocking tube(HBT)was used to process the specimens that were positive by the colloidal gold method, and then the test was performed again. Results: (1)Among the 8 678 patient specimens submitted for examination, 8 677 were negative for SARS-CoV-2 nucleic acid detection, of which 25 cases(0.288%)were positive for SARS-CoV-2 IgM antibody detected by colloidal gold method, SARS-CoV-2 -5 patients(0.058%)with IgG antibody positive, and 0 patients with SARS-CoV-2 IgM and SARS-CoV-2 IgG antibodies at the same time.(2)Thirty patients who tested positive for SARS-CoV-2 antibody by colloidal gold method had no epidemiological history, and at the same time, the results of SARS-CoV-2 nucleic acid test were negative. The possibility of SARS-CoV-2 infection or previous SARS-CoV-2 infection was ruled out. Among them, 10 patients with positive SARS-CoV-2 IgM antibody were still positive after more than 2 times of dynamic monitoring, and the remaining 20 patients were not clinically active monitored.(3)The false positive rate of SARS-CoV-2 IgM antibody detected by colloidal gold method was 0.288%, and the false positive rate of SARS-CoV-2 IgG antibody was 0.058%. After HBT was used to process the SARS-CoV-2 colloidal gold-positive specimens, among the 25 SARS-CoV-2 IgM antibody-positive specimens, except for 1 case which was still positive, the other results became negative;the results of 5 specimens with positive SARS-CoV-2 IgG antibody were still positive.(4)The rheumatoid factor, complement and immunoglobulin levels of 30 SARS-CoV-2 antibody-positive patients were within the normal reference range. Conclusion: The SARS-CoV-2 antibody test results may have false positives. The majority of SARS-CoV-2 IgM antibody positives are caused by heterophilic antibody interference. Those with SARS-CoV-2 IgG antibody positive may also have other potentially unknown interference factors. Heterophile antibodies interfere with colloidal gold methods more than chemiluminescence methods. Therefore, non-suspected patients or confirmed patients should not be tested for SARS-CoV-2 antibody to guide clinical practice, and the interference factors of the experiment must be considered. Copyright © 2021 by the Chinese General Practice.