Policymakers often lack reliable information on the long-term impacts of policy interventions in early childhood. Few studies attempt or are able to estimate effects on outcomes decades subsequent to an intervention. Even randomized controlled trials examining a single intervention generally are uninformative as to whether one policy administered to the same target population is more or less effective than another, whether the timing of interventions matters, or whether multiple interventions—concurrent or sequential—have notably bigger impacts than single ones. This paper assesses the value of using microsimulation methods to fill in knowledge gaps related to the long-term benefits of early interventions.  It also discusses the limitations inherent in these models.  The relevant issues are illustrated using the Brookings Social Genome Model, a new microsimulation model intended to flesh out the processes behind upward and downward mobility.  We provide details of the model and present results comparing the model’s estimates of the long-term effects of several interventions in early childhood against the findings from actual randomized controlled trials.  The results are suggestive of the strengths and weaknesses of the model and of microsimulation methods in general.