<![CDATA[Bullying in faith-based residential schools remains understudied, particularly where students share classrooms, dormitories, worship spaces, and daily supervision. This cross-sectional multisite survey examined the overlap between peer victimization and perpetration and assessed their associations with perceived anti-bullying policy implementation quality, safety, and trust in reporting mechanisms. Institution-mediated nonprobability sampling recruited 112 respondents from 11 modern pesantren across Jakarta, Bekasi, and South Tangerang in Greater Jakarta. Five reflective constructs measured through a 38-item questionnaire were analyzed using partial least squares structural equation modeling. The measurement model demonstrated acceptable internal consistency and convergent validity for victimization, perpetration, implementation quality, and reporting trust. However, the safety construct did not meet the recommended reliability and validity criteria and should therefore be interpreted cautiously. The structural model identified a positive association between victimization and perpetration, supporting the presence of victimization–perpetration overlap within the sample. Associations involving perceived implementation quality, safety, and reporting trust, including the proposed indirect relationships, were not supported. Given the cross-sectional design and nonprobability sample, these findings neither establish temporal ordering nor demonstrate policy effectiveness. Safeguarding systems in pesantren should identify students occupying both bullying roles, provide confidential reporting channels beyond immediate hierarchies, and monitor implementation fidelity. Future research should revalidate the safety measure and employ longitudinal or intervention designs to clarify directionality and assess safeguarding effectiveness.]]>
