This study evaluated the Survey of Income Program Participation (SIPP) as a source of data on children. The evaluation sought to determine whether SIPP provides reasonable estimates of the child population, by comparing estimates derived from SIPP with estimates derived from the Current Population Survey (CPS). Specifically, children living in families receiving Aid to Families with Dependent Children (AFDC), children living in families that are poor but not receiving AFDC, and children living in near-poor families were compared with children living in non-poor families and with all children in the United States; comparisons were made on basic demographic variables and on selected characteristics of the children's parents. With only a few exceptions, the estimates from SIPP were similar to those from the CPS, in spite of the SIPP smaller sample size. For some types of questions--specifically those related to employment, income, and Food Stamp receipt--data from SIPP may be better than those available from CPS because of the shorter recall period within SIPP for these questions. In addition to the comparison, a profile was created of children in the United States who are at risk for adverse outcomes because of living in families receiving welfare, living in or near poverty, or living with a mother who began childbearing when a teenager. (Contains 11 references and 8 tables.) (KA)