Based on the assumption that basic science is the crucial building block for technological and biomedical progress, this paper examines the relevance for public policy of basic demographic and behavioral sciences research on children and families. The characteristics of basic research as they apply to policy making are explored. First, basic research draws the attention of policy makers to problems, such as the relationship of teen motherhood to AFDC entry, the negative outcomes associated with teen childbearing, poverty, and divorce. Second, basic research can push policy makers away from issues that are not problems, such as expected but unsubstantiated problems for children of employed mothers. Third, basic research can help policy makers distinguish causal from correlated factors, such as determining that some negative outcomes of teen motherhood are due to the timing of the birth, but most of the negative outcomes are due to antecedent disadvantages, such as poverty. Fourth, basic research can contribute to program and policy evaluation by helping analysts develop models of behavior and providing measures and methods needed to conduct rigorous studies. The paper finds that with few exceptions, policy evaluation studies are not built on basic research; interventions are not designed with theoretical or model references, and evaluations are poorly designed. The paper also describes an evaluation study of the impact on children of the 1988 Family Support Act, which required that mothers of preschool children participate in education, job training, or work. The paper suggests that an examination of basic research guided this study's hypotheses and choice of methods and measures, and thereby will result in a much more informative study. (KB)