Between 1962 and 2005, the rate of female participation in the French labour force rose from 45.8% to 63.8%, but remained negatively correlated with the number of children. To what extent does an additional child reduce mothers’ participation rate? The relationship between number of children and participation is complex, for fertility and work decisions have common determinants and influence each other. It is therefore hard to say a priori if the choice of working or not is a cause or consequence of the fact of having a certain number of children. To test the existence of a negative causal relationship between number of children and mothers’ labour supply, we use instrumental variables, i.e., variables that have only an indirect effect on women’s labour-force participation, via their influence on the number of children. More specifically, the gender distribution of the two eldest children and having twins at second birth are two random sources of exogenous variation in fertility. And having the two eldest children of the same gender or twins at second birth increases the probability of having more than two children—in which case, mothers’ participation rate is reduced. The two variables allow us to estimate the causal influence of having more than two children on mothers’ participation rate. The results show that having more than two children reduces the probability of mothers’ participation by about 20 points and, when they are in employment, it reduces weekly working time by two hours. The negative impact on mothers’ participation may be proportional to the weakness of the mothers’ job and earnings prospects in the labour market and to their need to arrange for child care. This effect is particularly significant for mothers with low educational attainment, but persists when the children grow up and does not vary by size of locality of residence.