Poor infant and young child feeding (IYCF) practices are major determinants of chronic malnutrition. The main objective of this study was to assess the impact of a nutrition education (NE) programme aimed at promoting improved IYCF behaviours in combination with an agriculture intervention on children's dietary diversity and nutritional status. From 2012 to 2014, a cluster randomised trial was rolled out in Cambodia in the context of an agriculture and nutrition project of the FAO of the UN. The cross-sectional baseline study was carried out in sixteen pre-selected communes in 2012. Restricted randomisation allotted the communes to either intervention (NE and agriculture intervention) or comparison arms (agriculture intervention only). The impact survey was conducted as a census in all FAO project villages in 2014. Caregivers of children aged 0-23 months were interviewed using standardised questions on socio-economic status and dietary diversity (24-h recall). Anthropometric measurements were taken. A difference-in-differences model was applied. The sample comprised 743 households with children ≥6 months of age at baseline and 921 at impact. After 1 year of NE, 69 % of the intervention households reported to have participated in the NE. Estimated mean child dietary diversity was significantly different at impact between comparison and intervention (3·6 and 3·9, respectively). In particular, the consumption of pro-vitamin A-rich foods and other fruits and vegetables increased. No treatment effects on height-for-age Z-scores could be shown. NE led to improvements in children's diets. For effects on growth, it is assumed that longer NE activities are required to achieve sustainable behaviour change of age-appropriate infant feeding.