This paper investigates the effect of sibship size and birth order on educational attainment, for the United States and the Netherlands. An instrumental variables approach is used to identify the effect of sibship size. Instruments for the number of children are twins at last birth and the sex mix of the first two children. The effect of birth order is identified, by examining the relation with years of education for different family sizes separately; this avoids the problem that estimated effects confound birth order with family size. No significant effect of the number of children on educational attainment of the oldest child is found. Birth order has a significant negative effect. This negative effect does not differ between children from higher or lower educated parents. Also the age gap between children does not affect the effect of birth order, or the educational attainments of the children. These last two results suggest that competition between siblings for scarce parental time and resources is not an important cause of the birth order effects.