["Journal of Child Psychology and Psychiatry, EarlyView. ", "\n\nBackground\nGenetic differences are robustly associated with educational outcomes, but how they become linked is poorly understood. A plausible hypothesis, yet to be thoroughly empirically tested, is that the school environment mediates the association.\n\n\nMethods\nUsing structural equation models, we tested whether the student–teacher relationship at age 5 (teacher‐reported Student–Teacher Relationship Scale) mediated the association between children's education‐linked genetic propensities (PGSedu) and mother‐rated reading and math performance at ages 6–8. We performed these analyses in 63,032 children from the Norwegian Mother, Father and Child Cohort Study, a longitudinal, population‐based, pregnancy cohort.\n\n\nResults\nHigher PGSedu were significantly associated with lower student–teacher conflict (β = −0.08, p < .001) and positive teacher responses (β = 0.07, p = .001), but not student–teacher closeness (β = 0.01, p = .616). Associations between PGSedu and educational performance were significantly partially mediated via conflict (β = 0.01, p < .001) and positive teacher responses (β = 0.01, p = .019) but not closeness (β = 1 × 10−3, p = .619).\n\n\nConclusions\nStudent–teacher conflict and positive teacher responses may be mechanisms involved in the cascade of pathways linking children's genetics to their educational outcomes.\n\n"]