There is increasing research interest as to whether Dialectical Behaviour Therapy (DBT), specifically the group skills training component, is an effective early intervention approach when delivered universally in schools. The current study aimed to provide the first large-scale examination of a universal DBT-based intervention and to also determine the extent to which home practice of DBT skills predicted changes in social and emotional outcomes over time. A non-randomised controlled trial design was employed whereby 1071 participants (51.30% Male; M age = 13.48 years) completed either an adapted eight-session DBT skills-training intervention (‘WISE Teens) (n = 563) or class-as-per-usual (n = 508). On average, the ‘WISE Teens’ intervention did not improve outcomes with significant deteriorations or null effects observed across outcomes relative to class-as-per-usual immediately post-intervention. The largest deteriorations were observed for depressive and anxiety symptoms. Applying Bonferroni corrections, most group differences in outcomes had dissipated at follow-up, however, ‘WISE Teens’ participants continued to report significantly poorer quality of parent-child relationships relative to control (d = .16 for mother (95% CI = 0.01, 0.31); d = 0.17 for father (95% CI = 0.02, 0.33). While home practice was modest on average, further exploratory analyses nevertheless revealed that greater home practice was generally associated with more positive outcomes both immediately post-intervention and at follow-up. Based on these findings, the DBT-based ‘WISE Teens’ intervention is not recommended in its current format for universal dissemination amongst early adolescents in schools. Further research is needed to determine how to improve engagement and feasibility of delivery of DBT-based interventions universally in this context.