There has been a movement toward public policy that directs funding for programs and practice with evidence of effectiveness in achieving outcomes. For example, legislation authorizing the Maternal, Infant, and Early Childhood Home Visiting Program requires that most of the $1.5 billion dollars over the program's five years be spent on models that have evidence they improve outcomes. Evidence-based policy involves integrating the best scientific knowledge, clinical experience, and input from clients to choose the most effective course of action for a problem. Interest in applying these concepts to social policy comes, in part, from increasing pressure by federal, state, and local funders to demonstrate improved outcomes for children and families from social expenditures. There has also been a rapid increase in federally sponsored systematic reviews that support states' and communities' selection of programs or practices with evidence of effectiveness. However, most evidence is from small-scale, tightly controlled efficacy trials in a few communities, so it's unclear whether programs will show the same impacts in different communities with different populations. An emerging body of research (summarized here) is pointing to the importance of planning for and carefully structuring the transition from small and tightly controlled demonstration projects to large-scale implementation in communities. (author abstract)