A serosurvey of Hepatitis B infection markers was conducted in two orphanages that adhered to Hepatitis B vaccination policy. In spite of comparable sizes (80-90 children per facility), housing conditions and infection control practices, the level of HbsAg endemicity was different in each unit in direct relation with the mean age of the children. The prevalence of HbsAg carriers and the interval spent in collectivity strongly affect the seroconversion rate after HB vaccination. Other elements that can explain the low seroconversion rate were: the proportion on fully vaccinated children, the number of vaccine administered doses and the delayed age at which childhood immunization schedule was initiated. In order to increase the protective antibody response, booster doses were administered to a limited number of nonseroconvertors or to children with a nonprotective level of anti-HBs antibody (< 10 UI). This intervention provides evidence of prompt rising in antibody titers, comparable with titers found in children with wild infection.