Children in low-resource settings often consume microbially contaminated food, risking their health. We evaluated the impact of a food hygiene intervention on complementary food contamination in Bangladesh. A 3-year homestead food production intervention was complemented by an 8-month behavior change module to improve household food hygiene practices and evaluated in a cluster-randomized controlled trial, including a dedicated study measuring outcomes along the hygiene pathway to intestinal health. We used multilevel regression to assess the intervention's impact on microbial food contamination as well as food hygiene knowledge (n = 518) and reported practices (n = 531) among mothers of children 6-23 months of age. Complementary food samples were collected from 342 households with children 6-18 months of age and tested for Escherichia coli. Overall, 46% of food samples were contaminated with E. coli (43% intervention, 51% control), and there was no evidence that the intervention reduced food contamination (odds ratio: 0.65, 95% CI: 0.35-1.21). A higher proportion of intervention mothers could name all key food hygiene practices (23% intervention versus 1% control), had access to a basic handwashing station near the kitchen (24% versus 14%), reported washing hands before food preparation and child feeding (21% versus 8%), reported washing and storing feeding utensils safely (61% versus 49%), and reported preparing food fresh or reheating stored food (88% versus 79%) compared with control mothers. The intervention thus improved knowledge and reported food hygiene practices among mothers, but this improvement did not result in a substantial reduction of complementary food contamination.