In this paper we utilize a model of household investments in the development of children to explore the impact of various transfer policies on the distribution of child outcomes. We develop a cost criterion that can be used to compare the cost effectiveness of unrestricted, restricted, and conditional cash transfer systems, and find that an optimally chosen conditional cash transfer program is the most cost efficient way to attain any given gain in average child quality. We explore several design elements for the conditional cash transfer system and discuss the role of production function uncertainty and measurement error.