Family characteristics influence children's behavioral development, but so do variations in schools, neighborhoods, and communities. We documented extrafamilial environmental effects by fitting maximum likelihood models to questionnaire data collected from double dyads consisting of twins and their classmate controls. The classmate controls in each double dyad were genetic strangers living in separate households, but they shared school, neighborhood, and community environments with their yoked twin pair and with one another. At ages 11 to 12, the control classmates showed significant similarities in religious practices and smoking and drinking patterns, demonstrating that environmental influences outside the family affect children's behavioral development. Familial self-selection of residential neighborhoods may have contributed to these results, but direct effects of variation across communities, neighborhoods, and schools cannot be dismissed, and such effects warrant further study.