Immense expenditures support programs to prevent malnutrition, a global disease burden impairing performance, health and survival. Most programs are funded for 3-5 years and expected to obtain other resources for permanent and self-sustaining continuation. This thesis examines sustainability in 3 community-based child nutrition programs in the highland regions of Peru. The study aims were to: (1) examine different forms of program logic and elucidate program impact pathways (PIPs), (2) determine types and degrees of program sustainability, and (3) elucidate contextual factors that influence sustainability. We analyzed program documents, observed program activities in action, and conducted 20 semi-structured interviews with national, regional and local staff, in order to construct PIPs of key activities and examine them with program logic models. PIPs were a useful tool for mapping causal connections required for impact. Given the partial conceptual and operational knowledge among staff, communication across operational levels may lead to better understanding of causal mechanisms. Data collection for the sustainability assessment was conducted 1-4 years after project termination, and included 103 interviews with implementers and 28 focus groups with mothers in 28 communities, across the 3 different programs. To assess the degree of sustained activities, we adapted Pluye and colleagues' (2004) program sustainability framework based on organizational theory and operationalized characteristics of routinization (resources, adaptations, values and rules) and standardization (institutional standards). We found that the initial programs had disintegrated in all communities. However, a few activities continued in 9 communities at weak or medium sustainability levels, demonstrated by non-routinized or routinized activities without standardization. To determine factors associated with sustainability, we adapted Shediac-Rizkallah and Bone's (1998) framework for conceptualizing program sustainability. We identified common influential factors related to the initial program (broader community participation, positive perception of program impact, and intentional actions at exit), organizational factors (integration, external coordination, higher skills level and training, positive perceived value of work, strong work motivation, and champions for child nutrition), and community factors (perception of problem, integration, valuing of child nutrition, and champions). This research provides methods useful for evaluating sustainability and for potentially improving program design and implementation for sustainability.