Two different methods for preventing the binding of cross-reacting antibodies to the genus-reactive chlamydial lipopolysaccharide (LPS) were used to improve the specificity of an enzyme immunoassay for the determination of antibodies to Chlamydia trachomatis. Coated elementary bodies were treated with either sodium periodate, to oxidize the antigenic sites of the LPS, or Triton X-100, to extract the LPS. By using these new enzyme immunoassays, the standard enzymeimmunoassay, and the whole inclusion fluorescence (WIF) assay, antibodies to C. trachomatis were determined in sera from different groups of patients and controls. Paired serum samples from patients with culture-proven urogenital C. trachomatis infections showed similar responses in all three assays. Paired serum samples from patients with Chlamydia psittaci infections show,ed similar responses in the WIF assay and the standard enzyme immunoassay, whereas significantly reduced titers were obtained in the enzyme immunoassays with treated antigen, especially in the convalescent-phase serum samples. Serum samples from patients with symptoms suggestive of infection with C. trachomatis, pregnant women, and blood donors were evaluated by all three types of assays. Eighty percent of the significant reductions in immunoglobulin G (IgG), IgA, and IgM titers were observed in sera,vith WIF assay titers in the lower classes (IgG, 1:less than or equal to 256; IgA 1:less than or equal to 32; IgM, 1:less than or equal to 16). From these results we conclude that oxidation of the antigen by sodium periodate is a simple and effective method of producing an enzyme immunoassay, with enhanced specificity that could be useful for diagnostic purposes and seroepidemiological studies.