The Maternal Infant Health Program (MIHP) is Michigan?s largest evidence-based home visiting program for Medicaid-eligible pregnant women and infants. The purpose of MIHP is to promote healthy pregnancies, positive birth outcomes, and healthy infant growth and development, with the long-term goal of reducing infant mortality and morbidity. The program serves over 15,000 women and over 20,000 infants annually. Gaining a better understanding of its effectiveness is critically important. However, rigorously assessing the impact of MIHP is difficult because the program is available to all Medicaid-eligible women, so women cannot be randomly assigned to services. For this reason, prior research has used quasi-experimental analyses to assess the impact of the program. Unfortunately, these analyses rely on matching women on observable characteristics, but it is likely that unobservable characteristics contribute both to a woman?s decision to participate in MIHP and also to health outcomes. We propose using a randomized-encouragement design to more clearly identify the causal impact of MIHP on key health outcomes.