Bullying is common among school students, and some victims hold self-blaming attributions, exhibit low self-esteem, and do not seek social support. A wait-list control pre-/post-test experimental design, with random allocation, was used to assess the effects of a novel cross-age teaching of social issues (CATS) intervention on the latter 3 variables among peer-identified victims of bullying (N = 41, mean age = 14.5 years). In small cooperative groups of classmates, participants designed and delivered a lesson to younger students that informed them that bullies not victims are in the wrong, victims have no reason to feel bad about themselves, and that seeking help can be beneficial. CATS led to a significant improvement on all 3 dependent variables with mostly large effect sizes; these positive effects were even stronger with a bigger dose of intervention (6 hr vs. 4 hr), and changes in self-blame, and separately changes in self-esteem, mediated the positive effect of the intervention on help-seeking. The theoretical and practical implications of these results were discussed, especially in terms of supporting a highly vulnerable subgroup of adolescents.