Background: Syphilis remains a curable bacterial sexually transmitted infection (STI), with a global burden of 7.1 million new cases during 2020. Serological tests for syphilis require equipment unavailable in many health centers; however, point-of-care testing has allowed rapid screening using finger-prick whole blood samples. A further improvement could be the use of oral fluid. We evaluated the performance of treponemal antibody point-of-care testing for syphilis screening using oral fluid samples. Methods: Individuals who routinely attended STI clinics in Lima, Peru, were recruited. We collected oral fluid using the OraSure collection device (OraSure Technologies Inc., USA) according to manufacturer's instructions and serum from each participant. Oral fluid and serum were analyzed using the SD Bioline Syphilis 3.0 rapid test (Standard Diagnostics Inc., Korea). Oral fluid was processed using 30µL + 2 drops of sample diluent, while serum was processed using 10µL + 4 drops of sample diluent; serum were also tested using RPR (RPR slide test, Wiener Laboratorios SAIC, Argentina) and TPPA (Serodia, Fujirebio Diagnostics Inc., Japan). We assessed oral fluid rapid test overall percent agreement, sensitivity, and specificity against the following comparators: (1) serum SD Bioline Syphilis, (2) serum TPPA, and (3) both TPPA and RPR reactive or both TPPA and RPR non-reactive. Results: Among 324 participants, median age was 31.5 years (IQR: 11.5) and 71.6% were male. Nearly half (51.3%) reported prior syphilis and 30.6% were living with HIV. The overall percent agreement, sensitivity, and specificity of oral fluid was 71.0% (95% CI: 65.7% -- 75.9%), 78.2% (71.3% -- 83.8%), and 63.5% (55.8% -- 70.6%), respectively, versus serum SD Bioline Syphilis; similar results were obtained versus serum TPPA. The overall percent agreement and sensitivity increased when using the third comparator (74.6% [68.8% -- 79.8%] and 86.6% [78.9% -- 92.3%], respectively). When limiting the sample to TPPA-reactive with RPR titer ≥ 1:8, the sensitivity increased to 88.1% (75.0% -- 94.8%), but the percent agreement decreased to 70.4% (63.3% -- 76.9%). Conclusions: We observed a good performance of the rapid treponemal test using oral fluid. Further investigations are needed to improve the specificity of oral fluid as a potential sample for accurate syphilis screening. [ABSTRACT FROM AUTHOR]