In light of recent data suggesting adverse health effects at blood lead levels (PbB) <10 microg/dl, lowering the current definition of elevated blood lead (>or=10 microg/dl) has been recommended. To ascertain the population level impact of such a change, we calculated the prevalence of PbB >or=5 microg/dl in 1-21-year-old population in the United States. Furthermore, we characterized changes in PbB between 1988-1994 and 1999-2002. We analyzed data from the National Health and Nutrition Examination Survey (NHANES) III (n=10,755) and NHANES 1999-2002 (n=8013). In 1999-2002, about 91.7% of study children had detectable levels of lead in the blood. Among them, 7.3%, 2.8%, and 1.0% children and adolescents aged 1-5, 6-11, and 12-21 years, respectively, had PbB between 5 and 9.9 microg/dl. This number translates to approximately 2.4 million individuals. Between 1988-1994 and 1999-2002, the geometric mean PbB declined from 2.88 to 1.94 microg/dl in children 1-5 years, 1.80 to 1.36 microg/dl in children 6-11 years, and 1.24-1.02 microg/dl in children and adolescents 12-21 years of age. Also, the prevalence of PbB >or=5 microg/dl declined from 25.7% to 8.8%, 12.8% to 3.0%, and 7.5% to 1.2% in these age groups, respectively.A substantial proportion of children may be at risk for adverse health effects from lead exposure below 10 microg/dl and a large number of children will be classified as having elevated PbB if 5 microg/dl is considered the threshold. Significant public health resources will have to be mobilized for intervention, screening, and case management of these children.