The high maternal HBV DNA level is the most important factor contributing to HBV perinatal transmission. This study is to explore whether HBsAg can be used as a surrogate marker of serum HBV DNA for HBsAg-positive pregnant women.A total of 975 HBsAg-positive pregnant women and their infants were enrolled in this study. All infants received three doses of a yeast-derived recombinant Hepatitis B vaccine at 0, 1 and 6 months. They were also given Hepatitis B immunoglobulin (HBIG) at birth. HBsAg and HBeAg were determined using Abbott Architect assays while serum HBV DNA level was detected by the Abbott Real Time HBV DNA assay.Of the 975 subjects, 367 (37.6%) were HBeAg-positive and 608 (62.4%) were HBeAg-negative. Among the HBeAg-positive group, the samples with HBV DNA levels of ≥7.0 logIU/mL were 76.6% (281/367), and it was only 0.7% (4/608) for the HBeAg-negative group. HBV DNA level was positively correlated with HBsAg in HBeAg-positive group (r=0.786, p<0.001) but not in HBeAg-negative group (r=0.022, p=0.593). Among HBeAg-positive group, the area under the receiver-operator curve (ROC) of HBsAg titer for high HBV DNA level (≥7.0 logIU/mL) was 0.961 (95% CI, 0.940-0.983, p<0.001). The optimum cut-off point HBsAg titer above 4.1 logIU/mL had a sensitivity of 85.1%, specificity of 96.5%, and accuracy of 87.5% to predict HBV DNA levels of ≥7.0 logIU/mL. Of 367 infants born to mothers with HBeAg-positive, perinatal transmission was detected in 24 infants (6.5%, 24/367). Their mothers all had serum HBV DNA levels of ≥7.0 logIU/mL, 23 (95.8%) had HBsAg titers of ≥4.1 logIU/mL and the other mother had HBsAg titer of 3.9 logIU/mL. Of 608 infants born to mothers with HBeAg-negative, only one (0.2%, 1/608) became HBsAg-positive at the age of 7 months, and the mother of the infant had serum HBV DNA level of 3.4 logIU/mL and HBsAg titer of 1.8 logIU/mL, respectively.Serum HBsAg titer may be used as a surrogate marker of serum HBV DNA for HBeAg-positive pregnant women.