Maternal depressive symptoms are a robust risk factor for poor cognitive outcomes in children, yet the role of gene-environment interplay in this association is not well understood. The objective of this study was to evaluate gene-environment interaction in the association between maternal depressive symptoms and children's cognitive school readiness. Data come from a population-based birth cohort of 538 twin pairs. Maternal depressive symptoms were self-reported (Centre for Epidemiologic Studies Depression Scale) when children were aged 6 and 18 months (a mean score was used). Children's cognitive school readiness was assessed using the Lollipop Test when children were aged 5 years. Analyses were conducted with structural equation modeling. Maternal depressive symptoms were correlated with children's cognitive school readiness (r = -0.10). Shared environmental factors explained most of the variance in children's cognitive school readiness (52%). The remaining variance was accounted for by genetic (30%) and nonshared environmental factors (18%). As the level of maternal depressive symptoms increased, the relative contribution of nonshared environmental factors to the variance in children's cognitive school readiness increased (0.14 [95% CI: 0.04 to 0.24]), whereas the relative contribution of genetic factors decreased (-0.28 [-0.64 to 0.08]). In contexts of elevated maternal depressive symptoms, environmental - and potentially modifiable - factors may be especially important for shaping children's cognitive outcomes. This suggests that interventions to improve the early childhood environment of children exposed to maternal depressive symptoms may improve their cognitive outcomes.< Réduire