PurposeThis study aimed to investigate the effect of exercise intervention, and analyze exercise intervention strategies for adolescent depression through a meta-analysis of RCTs.MethodsAccordance to PRISMA guidelines, PubMed, Medline, EBSCO, Web of Science, SPORTDiscus, PsycINFO, ProQuest, and CNKI were searched for eligible records. Peer-reviewed studies were included if they met the following criteria: population (mean age of 10–18 years), intervention (physical activity, sport, or exercise), and outcomes (depression, adherence, ITT, dropout, adverse events, follow-up report). The protocol of this systematic review was registered in PROSPERO (CRD42022321683). Effect sizes calculations and methodological quality of exercise intervention (TESTEX scale) were carried out. The certainty of evidence was assessed by GRADE framework.ResultsThirteen randomized controlled trials were eligible for this review, which comprised a total of 433 adolescents. Compared with the control treatment, the effect of exercise on adolescent depression was moderate (SMD = −0.65, 95%CI: −1.03 to −0.27, p < 0.01). Heterogeneity was substantial (T2 = 0.30, I2 = 67%, p < 0.01). The moderating effect analysis showed that exercise intervention characteristics (organization form, exercise frequency, exercise intensity, exercise type, and single exercise session duration) of included studies varied greatly revealing multiple factors that may impact the antidepressant effect of exercise on adolescent depression (I2 > 50%, p < 0.05). Three studies show that the positive effect of exercise on reducing depression in adolescents remained 40 weeks after the intervention. Moreover, owing to the included studies contained methodological limitations, the certainty of evidence was reduced to moderate level.ConclusionThis study shows that exercise intervention has a moderate and sustained positive effect on adolescent depression. Our results recommended that adolescents with depression undertake moderate to high intensity group mixed exercise for more than 12 weeks, 20 to 60 min/time, more than 3 times/week. Additionally, our study also shows that the antidepressant effects remained for a long time after the end of exercise interventions. However, following the GRADE framework, we rated the certainty of evidence the primary meta-analysis as moderate evidence due to some limitations of included studies. Therefore, rigorous studies are still needed to verify the results.Systematic review registration[https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=321683], identifier [CRD42022321683].