BACKGROUND: Homestead food production programs have the potential to improve maternal and child health and nutrition through multiple pathways. OBJECTIVE: To evaluate the impact of a homestead food production program in Cambodia on household production and consumption of micronutrient-rich foods and on maternal and child health and nutrition (intake of micronutrient-rich foods, anthropometry, hemoglobin, and anemia prevalence); and to assess pathways of impact on maternal and child health and nutrition. METHODS: Two cross-sectional surveys (baseline and endline) were used to assess differences between intervention (n = 300) and control (n = 200) households using t-tests and chi-square tests. Using endline data and multivariate analyses, we examined the pathways of impact of the program on maternal and child health and nutrition. RESULTS: Intervention and control households were similar at baseline in sociodemographic characteristics, but more intervention households owned animals, earned income from homestead food production, and produced and consumed micronutrient-rich foods. At endline, some of these differences had widened; more intervention households produced and consumed more vegetables, had higher dietary diversity, and had a lower prevalence of fever among children under 5 years of age. In the intervention group, more children consumed more eggs and more mothers consumed micronutrient-rich food more frequently than in the control group. There were no other differences between the groups in maternal and child health and nutrition. Greater household production of fruits and vegetables was associated with greater household dietary diversity, which was associated with dietary diversity among mothers and children. Dietary diversity was not associated with other maternal and child health and nutrition outcomes. CONCLUSIONS: Cambodia's homestead food production program increased household production and consumption of micronutrient-rich foods and maternal and child intake (or frequency of intake) of some of these foods. Weaknesses in the evaluation design (e.g., lack of comparability between groups at baseline, failure to control for self-selection of households into the intervention, and collection of baseline and endline data during different seasons) prevent drawing firm conclusions about the program impacts. Analysis of impact pathways also shows that household-level benefits from the program did not translate into significant improvements in maternal and child health and nutrition. A careful redesign and rigorous assessment of the program using a program theory framework would help unleash its true potential to improve maternal and child health and nutrition outcomes.