This study aimed to analyze the associations of maternal characteristics and nutritional status of primary caregivers with the occurrence of stunting in children aged 6 – 59 months from rural agricultural households in Indonesia. This study used a sequential explanatory mixed method design. It started with a cross-sectional survey, followed by a qualitative study to explore community perceptions related to findings from the quantitative analysis. The survey involved 200 mother and child dyads. multiple logistic regression was carried out with the dependent variable being the incidence of stunting in children, and maternal characteristics as independent variables. Data analysis showed that WDDS and the maternal length of education in years were significant predictors for the incidence of stunting in children. The adjusted odds ratio (AOR) for WDDS was 0.78 (95% CI 0.63 - 0.98), while for the length of education of mothers was 1.22 (95% CI 1.01 - 1.48). We also conducted a qualitative study to explore food distribution and utilization within the household. Qualitative data analysis showed that, except for pregnant mothers, “good food” was prioritized the least for mothers. The persistent barriers for utilization of quality and diverse food are access to such food rather than knowledge of what is considered “good food for health”. Increasing access and consumption of diverse foods for all mothers beyond pregnancies is an important factor for preventing stunting in Indonesian children from rural agricultural households.