This study uses cross-sectional time-series data to examine the relationship between wealth and child labour and schooling in Pakistan and finds that wealth is crucial in determining a child's activities, but is far from being a sufficient condition to enrol a child in school. This is particularly the case for rural girls. Nonparametric analysis shows a universal increase in school enrolment for rural girls from 1998-2006 and this increase is independent of wealth. Multinomial logit regression further shows that wealth is insignificant in determining households' decisions about rural girls' activity. Thus, interventions to increase school enrolment should incorporate broadly targeted, demand-side interventions as well as supply-side interventions.