In the town of Netanya 9162 elementary schoolchildren (age 6-14 years) were tested for hepatitis B antigen (HBsAg) by the counter-electrophoresis method. The overall prevalence was 1.8% with a higher rate for males (2.2%) than for females (1.4%). Higher rates were observed among children born in the autumn and winter, particularly among the boys. This seasonal trend and male-female difference may partially be explained if infection occurs at circumcision. The ethnic distribution showed highest rates among children of Libyan and Yemenite origin (3.9% and 2.6%, respectively), and lowest rates among children of European and Israeli born parents (0.4% and 0.2%, respectively). Prevalence of HBsAg among children from large families and those from families living in poor and crowded quarters was very significantly higher than among children from smaller families living in richer areas and in less crowded conditions. These variables, which are closely associated with social class, remain significant after adjustment for ethnic origin and seem to account for a large proportion of the variation among the Israeli schoolchildren. The variation due to ethnicity is markedly reduced when adjusted for five other pertinent variables: age, sex, season of birth, crowding and family size. Assuming that prevalence of HBsAg in a population reflects the rates of infection (past or present), the results of this study strongly support the hypothesis that type B hepatitis in Israel behaves essentially as a contagious infection manifested primarily as a subclinical disease of early and mid-childhood.