OBJECTIVE To evaluate the efficacy of interruption of intrauterine infection of HBV with HBIG in pregnant women with positive HBeAg and HBsAg. METHODS A prospective randomized controlled trial was adopted. Sixty cases with positive HBeAg and HBsAg were coincident with the criteria of inclusion, and 8 cases were excluded. Fifty-two cases were analyzed (28 cases in trial group and 24 in control group). All cases in trial group received 200 IU HBIG intravenously every 4 wk for 3 times from the 28(th) wk. The cases of control group received placebo in the same way. All pregnant women were detected for HBeAg and HBV-DNA at the beginning of the trial and end of the trial (delivery). The cord blood of all newborns were collected for detecting HBeAg and HBV-DNA simultaneously. RESULTS For investigation of HBeAg of newborns in trial group, 6 of 28 cases of newborns had positive HBeAg, the HBeAg positive rate being 21.4%, the total rate of 95% CI being 8%-41%. In control group, 19 of 24 cases of newborns had positive HBeAg, HBeAg positive rate was 79.2%, the rate of 95%CI being 5%-93%. By statistical analysis, chi(2) = 17.26, P < 0.01, RR = 0.27, 95% CI (6.3 multiply 10(-6), 8.6 multiply 10(-5)). For investigation of HBV-DNA of newborns in trial group, 7 of 28 cases of newborns had positive HBV-DNA, HBV-DNA positive rate being 25%, the total rate of 95% CI being 11%-45%. In control group, 20 of 24 cases of newborns had positive HBV-DNA, HBV-DNA positive rate was 83.3%, the total rate of 95% CI being 63%-95%. By statistical analysis, chi(2) = 17.62, P < 0.01, RR = 0.30, 95% CI (1.5 multiply 10(-5), 1.7 multiply 10(-4). The results indicated that there was significant difference in HBeAg positive rate and HBV-DNA positive rate of newborns between the two groups. In trial group, 7 of 28 newborns had HBV-DNA positive, but the HBV-DNA load of newborns was lower than that of their mothers. In control group, 20 of 24 newborns still had HBV-DNA positive, and the HBV-DNA load of newborns was close to those of their mothers. Statistical analysis indicated that there was no significant difference in HBV-DNA load between postnatal women without HBIG intervention and their filial generations (T = 81.5, P > 0.1). CONCLUSIONS It is effective and safe to prevent intrauterine infection of HBV with HBIG from the 28(th) wk in pregnant women with positive HBeAg and HBsAg. In clinical application, those pregnant women with negative HBeAg and positive HBV-DNA also need to be interrupted by HBIG.