If maternal expressed emotion is an environmental risk factor for children's antisocial behavior problems, it should account for behavioral differences between siblings growing up in the same family even after genetic influences on children's behavior problems are taken into account. This hypothesis was tested in the Environmental Risk Longitudinal Twin Study with a nationally representative 1994-1995 birth cohort of twins. The authors interviewed the mothers of 565 five-year-old monozygotic (MZ) twin pairs and established which twin in each family received more negative emotional expression and which twin received more warmth. Within MZ pairs, the twin receiving more maternal negativity and less warmth had more antisocial behavior problems. Qualitative interviews were used to generate hypotheses about why mothers treat their children differently. The results suggest that maternal emotional attitudes toward children may play a causal role in the development of antisocial behavior and illustrate how genetically informative research can inform tests of socialization hypotheses.