This study examined the impact of contract farming on household food security in Kenya and Madagascar in the context of large-scale agricultural investments. An endogenous switching regression model was used to control for a possible selection bias due to unobserved factors. In general, the household dietary diversity score, food consumption score and months of adequate household food provisioning improved the food security of the households engaged in contract farming in Kenya. While in Madagascar, only the months of adequate household food provisioning had a positive impact. These results revealed that contract farming models do not always have a positive impact on food security. The effect on household food security could depend on the local context and crop under contract.