Seventy-eight infants born to HBsAg-positive women were randomly assigned to receive either the plasma-derived vaccine or 0.5 ml (10 micrograms HBsAg) yeast-derived recombinant hepatitis B vaccine within 24 hours of birth, simultaneously with hepatitis B hyperimmunoglobulin. In 67 infants who received the plasma-derived vaccine, one of the doses of 0.5 ml (25 micrograms HBsAg) was used randomly. In all of the infants, the second and third doses of both vaccines were given at one and two months of age, respectively. The booster doses were given at 12 month of age in all of the infants. These vaccinated infants were followed up until 13 months of age. There were differences in the seroconversion rates with different vaccines and doses. The recipients of the half-dose of plasma-derived vaccine showed lower seroconversion rates than the others, and the newborns in this group showed more seronegativity (13.2%) than the others (p < 0.05). The lowest anti-HBs geometric mean titers (GMTs) were obtained in newborns vaccinated with Hevac B 0.5 ml. Sixty percent of the anti-HBs GMTs in this group were under 100 mlU/ml. There were statistically significant differences between males and females in anti-HBs seronegativity rates, with males having lower anti-HBs GMTs than females. The difference was particularly significant among male newborns vaccinated with a half-dose of plasma-derived vaccine.