Persistent child poverty and awareness that noncustodial parents contribute to this problem by failing to pay child support has spurred interest in a national child-support assurance program (CSAP). We analyze several variations of a CSAP and find that the policies considered have a limited impact on child poverty. Moreover, the alternative which reduces child poverty the most does so at a high cost. Today, one in five children lives in poverty and over half of these children live in families with a parent living elsewhere. Yet, many of these children do not receive income from their noncustodial parent. Several recent proposals offered in Congress would establish a national or demonstration assurance program, whereby the government would pay a benefit to children when child support is not received. A child-support assurance demonstration project was also included in the final recommendations of the bipartisan National Commission on Children (1991). This paper examines the extent to which a national CSAP would reduce child poverty and at what cost. We use the Urban Institute's unique microsimulation instrument, the Transfer Income Model version 2 (TRIM2), to estimate the costs and effects of this policy using 1990 data. I. Designing a Child-Support Assurance Policy