["Journal of Child Psychology and Psychiatry, EarlyView. ", "\n\nBackground\nPast studies indicate significant associations between birth anthropometrics, such as relative birthweight or head circumference, and later cognitive performance. Using twin data, we investigated whether these associations are confounded by genetic and environmental factors. Additionally, we determined whether the association of within twin differences in birth anthropometrics and cognitive performance is nonlinear or moderated by factors such as zygosity or age at assessment.\n\n\nMethods\nLongitudinal data from over 2000 twin pairs from the German TwinLife study were analysed using the co‐twin control method. Cognitive performance was assessed at two waves, with median ages of 12 and 18 years, with z‐standardised scores based on wave and age. Differences between twins in birthweight and head circumference Z scores were calculated using Fenton's size at birth chart, based on health records at birth.\n\n\nResults\nWithin twin differences in relative birthweight were associated with differences in cognitive performance at Wave 1 (β = .08, p = .001) and did not differ by zygosity (β = .03, p = .560). However, this association was not significant when reducing the sample to twins with valid data at both waves (β = .05, p = .187). Testing for nonlinearity or for moderation by age at assessment did not improve model fit. Within twin differences in head circumference were not significantly associated with differences in cognitive performance in any analyses (smallest p = 0.365).\n\n\nConclusions\nAfter controlling for confounding due to genetics or the shared environment, conditions in the womb, as indexed by relative birthweight, is associated with later cognition. This association is not stronger in younger individuals.\n\n"]