This study tested 2 related hypotheses. The first holds that high co-rumination anticipates heightened internalizing problems. The second holds that positive relationships with friends exacerbate the risk for internalizing problems arising from co-rumination. A sample of MZ twins followed from birth (194 girls and 170 boys) completed (a) self-reports of friendship support, friendship negativity, and co-rumination with friends at age 12 and (b) measures of anxiety and depression at ages 12 and 13. Using a monozygotic twins-difference design, within-pair differences in co-rumination predicted increased within-pair differences in anxiety (but not depression), after removing the covariance between co-rumination and perceptions of friendship. In other words, the difference in co-rumination within each monozygotic twin pair predicted an increase in the difference in their anxiety levels, but not the difference in their depression levels. The discussion focuses on nonshared environmental influences, because the monozygotic twin-difference design eliminates the possibility that associations were driven by heritability or by shared environmental factors that underlie friendship experiences and internalizing problems.