Background Early screening for syphilis among pregnant women and the effective treatment of maternal syphilis is fundamental to prevent congenital syphilis (CS). Methods We obtained data from the Shenzhen Program for Prevention of CS (SPPCS) and estimated incidence rates of CS among infants born to syphilis-seropositive women treated with different regimens or untreated for maternal syphilis. Results A total of 4746 matched cases of syphilis-seropositive mothers and their infants were included for analyses, and 162 infants were diagnosed with CS, providing an overall incidence of 3.41% (95% confidence interval [CI], 2.91%-3.98%). Among infants born to syphilis-seropositive women who had syphilis and were adequately treated before pregnancy, the incidence was 0.22% (95% CI, .05%-.66%). There were 159 cases of CS occurring in 3519 infants born to women who were syphilis-seropositive during their pregnancies, for an incidence of 4.52% (95% CI, 3.84%-5.28%). The incidence of CS was 1.82%-11.90% lower among infants born to the women treated with early benzathine penicillin G (BPG) compared with those treated with late BPG (adjusted odds ratio [aOR], 8.06 [95% CI, 2.93-22.21]; P < .001), other antibiotics (aOR, 7.71 [95% CI, .86-69.28]; P = .068), or those untreated (aOR, 68.28 [95% CI, 29.64-157.28]; P < .001). The incidence rates were 0.22% (95% CI, .06%-.80%) and 0.59% (95% CI, .35%-1.02%) in infants born to women treated with 2 courses and 1 course of BPG, respectively, corresponding to a risk difference of 0.37% (aOR, 1.74; 95% CI, .37-8.26). Conclusions Treatment of syphilis-seropositive pregnant women with 1 course of intramuscular BPG before 28 gestational weeks is critical for prevention of CS.