Malnutrition and micronutrient deficiencies such as anemia have well-documented consequences on children’s health and schooling, cognitive functioning, and long term economic returns. This paper addresses the potential role of mother’s knowledge on reducing malnutrition and micronutrient deficiencies drawing on the social networks literature to better understand the role of knowledge provider characteristics and social network structure have on knowledge diffusion. A cluster randomized control trial was implemented to test an agricultural production and nutrition education program, targeted to mothers of children aged 3-12 months at baseline living across four districts in Gourma province, Burkina Faso. Villages were randomly assigned to either the control group or to one of two treatment groups based on knowledge provider characteristics. Both treatment groups participated in a two-year program that included homestead food production, aimed to increase production of high quality foods, and a nutrition information component, aimed to improve knowledge and adoption of optimal infant and young child feeding (IYCF) practices. The two treatment groups differed in the actors who delivered the nutrition information messages, either older women leaders (OWL villages) or village health committee members (HC villages). This paper investigates the role that the nutrition information delivery strategy and social networks play in diffusing nutrition information among young mothers. The diffusion strategy may affect the young mother’s confidence in and retention of new nutritional information, while social network structure may reinforce nutritional messages through peer effects. We find IYCF knowledge diffusion is higher in HC villages relative to OWL villages, while social network measures, degree and betweeness, have statistically significant, but smaller effects on certain categories of IYCF knowledge. The increased knowledge effects explain one of the causal pathways through which the intervention has an impact on childhood anemia (26% of a standard deviation) in the HC treatment, but not the OWL treatment.