This study aimed to test the effects of a 25-session version of the Growing Pro-Social (GPS-25) program over schemas and schema-related emotions in male young offenders. Participants included 123 youth aged between 14 and 19 years, placed in eight Portuguese detention facilities. Youth were allocated to receive GPS (n = 63) or treatment as usual (n = 60), and answered a self-report measure assessing schemas and schema-related emotions at baseline and posttreatment. Two-factor mixed multivariate analysis of variance (MANOVA; group change) and the Reliable Change Index (individual change) revealed nonsignificant differences between groups for the schema’s endorsement. Significant differences between groups were found for the schema-related emotions: Treatment participants presented lower scores and/or higher clinical improvements after GPS, when compared with controls. GPS-25 produced change at an emotional level but not in schema’s endorsement, suggesting that longer interventions should be tested in their capability to promote cognitive and emotional change in young offenders.