An investigation of whether and to what extent a peer-led model is able to counteract mechanisms underlying bullying in peer groups, seeking clarification of divergence in reported results on the efficacy of peer-led models. Two studies were carried out in Italy within a project tackling bullying and cyberbullying in secondary schools. In the first study (n= 386), concerning the first phase of the project, a significant decrease was found only for cyberbullying, most of all for male peer educators. For the second study (n= 375) the model was improved and significant effects were found for several participating groups (peer educators and the experimental classes), who exhibited a decrease in bullying, victimization, and cybervictimization. Results suggest that peer educators can act as agents of change in the broader context.