The existence of depression among children or adolescents can trigger a sequence of burdens on themselves, their families and even the whole society, which cause both physical and economic impacts. Our network meta-analysis (NMA) was aimed at comparing them with each other for evaluating the optimal psychosocial therapy to help children and adolescents with depression to improve their mental health. Based on several biomedical databases, a system of search strategies was conducted for searching randomized controlled trials (RCTs) which published from their inception on October, 1st 2018 without language restriction. We carried out an expression analysis for comparing the efficacy of various psychosocial therapies using Bayesian NMA. A battery of analyses and assessments, such as conventional meta-analysis and risk of bias, were performed concurrently. Only 32 of RCTs which involved 2677 participants were eventually included in our analyses from the 19,176 of initial citation screening. In addition, in terms of various valid assessment instruments, interpersonal psychotherapy [standard mean difference (SMD) = -1.38, Credible interval (CrI) - 2.5, - 0.20)], computer cognitive-behavioral therapy (SMD = -1.36, CrI - 2.59, - 0.14), cognitive-behavioral therapy (SMD = -1.16, CrI - 2.15, - 0.18), had significantly better effects than the named control group. All together, these results suggest that interpersonal psychotherapy might be the best approach to improve the depressive state among children and adolescents. This study may provide an excellent resource for future endeavors to utilize psychosocial interventions and may also serve as a springboard for creative undertakings as yet unknown.