Perinatal transmission of hepatitis B virus (HBV) from HBsAg carrier mothers who were HBeAg+, antiHBe+, or negative for both HBe markers, was interrupted using either 4 doses of vaccine, or one dose of hepatitis B immunoglobulin (HBIG) at birth, combined with 4 doses of vaccine. In those infants who received HBIG at birth, the antiHBs titre was significantly higher at 1 and 2 months old, but at 6, 9, and 18 months old, there was no significant difference. Among the infants of carrier mothers who did not display HBeAg (i.e., were antiHBe+, or negative for both HBe markers), a transient subclinical infection would have been expected in around 10% had there been no intervention. No evidence of such infection was detected, and no difference in outcome was found between the two treatment groups. Amongst infants born to HBeAg+ carrier mothers, infection occurred in 1 out of 8 who had received HBIG and vaccine, and in 3 of 8 who had received vaccine only. The difference in outcome was not statistically significant, but the numbers analysed were small. The infections which occurred in spite of prophylaxis may be attributable to in utero infection, poor response to vaccine by the infant, or to the mother having a particularly high HBV-DNA level. HBIG given at birth to infants of HBeAg+ carrier mothers may enhance the protection of infants who are destined to be poor responders to vaccine.