Contract farming has frequently been shown to increase the income of participating households. Whether contract farming increases other aspects of household welfare, however, remains unclear. Using a 1,200-household data set from Madagascar and the results of a contingent valuation experiment aimed at eliciting respondent's willingness to pay to participate in contract farming, we show that for the average household, participating in contract farming (i) reduces the duration of the hungry season experienced by the household by about 10 days and (ii) increases the likelihood that the household's hungry season will end by almost 20 percent in any given month. Further, we find that these effects are even more pronounced for households with a larger number of children and for households with a larger number of female children, who often bear a disproportionate share of the burden of food insecurity.