The performance of a latex agglutination test (Mediace TPLA) in the detection of anti-treponemal antibody was evaluated in comparison with chemical luminescence tests (LumipulsII-N and Architect TPAb) in 346 cases. Anti-treponemal antibody was further determined by immunochromatography and immunoblotting tests and additionally evaluated by a serological test for syphilis with lipoidal antigens. The total concordance rate between the latex agglutination test and chemical luminescence tests ranged from 96% to 97%: the positive concordance rate ranged from 96% to 97%, and the negative concordance rate, from 97% to 98%. The latex agglutination test showed two false positive cases, and each chemical luminescence test showed two false positive cases, respectively. In eight cases, only the latex agglutination test showed negative results; all specimens contained anti-treponemal antibodies. However, none of these was considered to be a false positive and each was treated as syphilis based on the results of confirmatory analysis with immunochromatography and immunoblotting tests and a serological test for syphilis. The discordant results in the latex agglutination test and chemical luminescence tests may be caused by the different antigenisity of each test. With detailed analysis of those sera treated as syphilis, each specimen was found to contain various antibodies against syphilitic antigens, suggesting that there was a different specificity of native and recombinant antigens. Based on the present results for the comparison between the latex agglutination test and chemical luminescence tests, it was considered that further investigation is necessary to clarify the anti-treponemal antibody profile of syphilis at the disease stage.