Out-of-wedlock childbearing has increased dramatically in the US during the last four decades. In 1960 about 6 percent of all children were born outside of marriage. Today, the figure is 33 percent. The numbers are even higher among racial and ethnic minorities. In 1999, 69 percent of African American children and 42 percent of Hispanic children were born to unmarried parents, as compared to only 22 percent of non-Hispanic whites. In response to these disturbing trends, Congress has made marriage a major issue in welfare reform. In 1996, the Personal Responsibility and Work Opportunity Act (PRWORA) offered financial rewards to states that reduced their non-marital births rates. More recently, as part of the reauthorization of PRWORA, the Bush administration proposes to spend up to $300 million dollars annually on demonstrations to promote "healthy marriages." The Administration also proposes to add the promotion of "father involvement" and "child wellbeing" to the objectives of PRWORA. In this paper we describe a demonstration designed to evaluate alternative strategies for promoting healthy marriages, father involvement, and child wellbeing among new unmarried parents. Regardless of the success of pro-marriage policies, a substantial proportion of unwed parents may choose to live apart; and policies that address the needs of these parents and their children may end up discouraging marriage. We discuss this trade-off in our proposal and incorporate a new strategy for resolving it. The demonstration would be evaluated by a random assignment experimental design and would include new services as well as changes in policies. Though the proposed demonstration has many components and would be expensive if enacted as national policy, it is no panacea. It does nothing to prevent unintended pregnancies or early childbearing. Still, we believe the intervention described in this chapter would lead to more stable unions, more father involvement, and better outcomes for children. Although the policy changes and new services we propose would be limited to unwed parents for the purpose of the demonstration, with one exception we envision them as being extended to a much broader population, including married parents and adults without children. To the extent that the demonstration induces unmarried parents to marry or become better parents, it also would reduce divorce and non-marital childbearing if provided to married couples and adults without children. Indeed, not extending benefits to married couples and childless adults might encourage divorce and unwed parenthood.