Are individual differences in attachment security inborn or shaped by the social environment? In infancy and early childhood, the evidence points to a substantial role of the environment, but a large twin study in early adolescence showed considerable heritability. Here we examined the twin heritability of attachment in middle childhood. We hypothesized that in middle childhood some heritability would emerge. Furthermore, we expected a role for cognitive and language abilities in explaining variance in attachment in middle childhood, partly related to the measurement of attachment, and we therefore examined associations with IQ.