OBJECTIVE To evaluate a treponemal enzyme-linked immunosorbent assay (ELISA) as an alternative screening test for syphilis in pregnant women. METHODS A cross-sectional study of diagnostic test accuracy was carried out in a large volume laboratory from a tertiary care center. A total of 416 serum samples, including 102 archived syphilis Treponema pallidum hemagglutination (TPHA)-positive samples and 314 samples from pregnant women, were used to determine the sensitivity and specificity of ELISA. All the samples were subjected to Venereal Disease Research Laboratory (VDRL), rapid plasma reagin (RPR), ELISA, and TPHA tests. Performance characteristics of VDRL, RPR, and ELISA were calculated with TPHA as a reference standard test. RESULTS VDRL and RPR exhibited higher false positivity of 10.5% and 9.6%, respectively, compared to 2.5% by ELISA. The sensitivity and specificity of ELISA were 98% and 97.5%, of VDRL were 71.6% and 89.5%, and of RPR were 73.5% and 90.5%, respectively. Moreover, ELISA had an excellent agreement (kappa=0.9) with TPHA compared to VDRL/RPR which had a moderate agreement (kappa=0.6) only. CONCLUSIONS ELISA has the potential to replace VDRL/RPR as a screening test for syphilis in centers that can perform ELISA, especially for antenatal screening.