BackgroundThere is evidence that birth weight is positively associated with education, but it remains unclear whether this association is explained by familial environmental factors, genetic factors or the intrauterine environment. We analysed the association between birth weight and educational years within twin pairs, which controls for genetic factors and the environment shared between co-twins.MethodsThe data were derived from nine twin cohorts in eight countries including 6116 complete twin pairs. The association between birth weight and educational attainment was analysed both between individuals and within pairs using linear regression analyses.ResultsIn between-individual analyses, birth weight was not associated with educational years. Within-pairs analyses revealed positive but modest associations for some sex, zygosity and birth year groups. The greatest association was found in dizygotic (DZ) men (0.65 educational years/kg birth weight, p=0.006); smaller effects of 0.3 educational years/kg birth weight were found within monozygotic (MZ) twins of both sexes and opposite-sex DZ twins. The magnitude of the associations differed by birth year in MZ women and opposite-sex DZ twins, showing a positive association in the 1915–1959 birth cohort but no association in the 1960–1984 birth cohort.ConclusionAlthough associations are weak and somewhat inconsistent, our results suggest that intrauterine environment may play a role when explaining the association between birth weight and educational attainment.