A series of bivariate probit regressions identifies the determinants of child labour and school attendance at the household level. Children from poor households and from households without a formally educated household head are more likely to be engaged in child labour and less likely to attend school than members of rich households and children living with an educated household head. This finding lends strong support to the hypothesis that poverty is the root cause of child labour. The paper concludes with recommendations for targeted cash transfers as a means to increase school attendance and reduce child labour.