BACKGROUND: Primary Epstein-Barr virus (EBV) infection affects the host differently according to when in life it is acquired. Understanding risk factors for infection could be important for disease prevention, and the age-specific prevalence of infection must be known to optimize use of a prophylactic vaccine. METHODS: Children 18 months to 19.9 years of age who had blood drawn for medical indications during an outpatient visit were eligible. Sera were tested for immunoglobulin G antibodies against EBV viral capsid antigen by enzyme immunoassay. Family demographic and socioeconomic data were obtained via scripted telephone questionnaires. RESULTS: Consent was given for 876 of 914 (96%) subjects approached. Sera were available for 782 of 876 (89%) subjects and demographic/socioeconomic data obtained for 705 (90%) of them. Antibody prevalence, adjusted for age and sex, was as follows: non-Hispanic blacks, 74%; Asians, 62%, multiracial children, 54%; Hispanics, 50%; and non-Hispanic whites, 26%. The pattern of increases in antibody prevalence with age differed significantly by race/ethnicity, and was most divergent in the 2 youngest age groups. Adjusted EBV antibody prevalence decreased with greater household education among non-Hispanic whites, but was not associated with any other socioeconomic factor. In 42 of 51 (82%) families with >1 child in the study, the siblings' EBV antibody status was concordant (bootstrap P < .001). CONCLUSIONS: Racial/ethnic differences in EBV antibody prevalence and concordance of antibody status among siblings prompt us to speculate that both genetics and family environment contribute to acquisition of EBV infection. The ideal age to give a prophylactic vaccine may differ according to race/ethnicity.