This study investigated how birth spacing affects children’s educational achievement using Panel Study of Family Dynamics (PSFD) data for 1999–2018. We converted the panel data into a structure of sibling pairs to estimate the effect of birth spacing on the educational achievements of both older and younger siblings. To account for the endogeneity of birth spacing, we used multiple births as the instrument. Our two-stage least squares (2SLS) estimates indicate that birth spacing positively affects the educational achievement of older siblings. A one-year increase in age difference raises the likelihood of children obtaining a college degree by approximately 2.7%, but spacing has no significant effect on younger siblings. Furthermore, by distinguishing the birth years of the samples, we found that the effects of birth spacing were in opposite directions in different generations. This suggests that the channels through which birth spacing affects educational outcomes vary significantly between early and modern societies.