The paper deals with child labour in developing countries. We address a problem that has recently drawn much attention at the international level, that is, how to invest in women’s rights to advance the rights of both women and children. We study the problem from a new perspective. In our theoretical model we assume that the child’s time is an extension of her/his mother’s time, and that she has to decide how to allocate it. We estimate two empirical specifications, both multinomial logit. The first one, in line with the standard approach in the literature, estimates a model of the probability of the different child’s states, conditional on her/his mother’s states. The second empirical specification, in line with our theoretical model, estimates the mother-child states jointly. Using a unique, rich and representative data survey for all Indian states and for urban and rural India (NFHS-2, 1998/9), we select our sample drawing information from the household data set and the women’s data set. Our results show that the presence of the mother in the family increases children welfare, in terms of educational opportunities and protection from work activities. All our results indicate that the mother tends to stay home and send her children to school the better is the father’s employment position and the wealthier is the family. However, we observe a perverse effect. If the mother works, since female job quality and wage levels are very low, also her children have a higher probability to work.