The objective of this thesis is to estimate causal effect of childbearing on women’slabor supply in Sweden. I follow the approach suggested by Angrist and Evans (1998)using parental preferences for a mixed child-gender composition as an exogenoussource of variation in women’s fertility. The results show that having an additionchild have a negative effect on women’s working hours. However, none of theseeffects are statistical significant and the value of F-statistic and partial-R2 are rathersmall, all suggest that the same-sex is very likely a weak instrument in Sweden.