To assess the clinical applicability of a semi-quantitative luciferase immunosorbent assay (LISA) for detecting antibodies against Treponema pallidum antigens TP0171 (TP15), TP0435 (TP17), and TP0574 (TP47) in diagnosing and monitoring syphilis. LISA for detection of anti-TP15, TP17, and TP47 antibodies were developed and evaluated for syphilis diagnosis using 261 serum samples (161 syphilis, 100 non-syphilis). Ninety serial serum samples from 6 syphilis rabbit models (3 treated, 3 untreated) and 110 paired serum samples from 55 syphilis patients were used to assess treatment effects by utilizing TRUST as a reference. Compared to TPPA, LISA-TP15, LISA-TP17, and LISA-TP47 showed a sensitivity of 91.9%, 96.9%, and 98.8%, specificity of 99%, 99%, and 98%, and AUC of 0.971, 0.992, and 0.995, respectively, in diagnosing syphilis. Strong correlations (rs = 0.89-0.93) with TPPA were observed. In serial serum samples from rabbit models, significant differences in the relative light unit (RLU) were observed between the treatment and control group for LISA-TP17 (days 31-51) and LISA-TP47 (day 41). In paired serum samples from syphilis patients, TRUST titres and the RLU of LISA-TP15, LISA-TP17, and LISA-TP47 decreased post-treatment (P < .001). When TRUST titres decreased by 0, 2, 4, or ≥8-folds, the RLU decreased by 17.53%, 31.34%, 48.62%, and 72.79% for LISA-TP15; 8.84%, 17.00%, 28.37%, and 50.57% for LISA-TP17; 22.25%, 29.79%, 51.75%, and 70.28% for LISA-TP47, respectively. Semi-quantitative LISA performs well for syphilis diagnosis while LISA-TP17 is more effective for monitoring syphilis treatment in rabbit models and clinical patients.