Motivation: Sub-Saharan African governments have subsidized farm inputs-fertilizer and seed especially-to increase food production by small-scale farmers to improve food security. A potential drawback of such schemes is that they may encourage farmers to put their children to work in the fields, harming their education. Purpose: Did the Malawi Farm Input Subsidy Programme that began in 2005/2006 increase child labour on the holdings of beneficiary smallholders? Methods and approach: The article analyses data from the Malawi Integrated Household Panel Survey to examine the effect of seed and fertilizer subsidies on child labour. The study employs a correlated-random-effects-control function regression, using district coupon allocation as an instrumental variable for coupons received by households. Findings: There was statistically significant evidence that the Farm Input Subsidy Programme (FISP) increased child labour. The effect, however, was relatively small. At the sample mean, it was estimated that the programme led to a 12 percentage point increase in the likelihood that children would work on the farm and that the children would work an additional 72 minutes a week on the fields. The FISP, however, did not affect the enrolment of children in school. Effects varied socially: children in male-headed, uneducated, and smallholder households were the most affected. Policy implications: Although the observed effects are not large, they are unwelcome. Two policy corrections could eliminate them. One, the award of subsidy coupons could be made conditional on children's school performance. Two, given that the effects barely applied in households where parents had been to school, agricultural training should stress the importance of children attending school and not working in the fields.