This paper examines the empirical evidence underlying Senegalese household decision making of whether to send children to work or to school. Using household survey data (Esam-I), we test the impact of household living standard and living arrangements, head of household background, adult activities, gender balance of power, household facilities and ethnic differences on children activities. We use a multinomial logit model to analyse the participation of children in three mutually exclusive activities: school, housework and work. The originality of this paper is to highlight the importance of opportunity costs of schooling (proxied by the presence of household facilities decreasing the needs for child housework) vis-à-vis the other alternatives. The paper also investigates the impact of female bargaining power on children activities. We find that children schooling – particularly of boys - increases when such a power becomes more important. Furthermore, estimates show that when households’ adult income increases, the probability for boys and girls to work decreases significantly while their enrolment in school increases, just as predicted by the Basu and Van’s luxury axiom. Estimates also indicate that adult background and cultural differences affect significantly the probability for children to be involved in one or another activity. Moreover, there is no evidence of a role of child-minders for older children : the presence of younger children does not affect older children activities