Background and AimsPeer‐led interventions may offer a beneficial approach in preventing substance use, but their impact has not yet been quantified. We conducted a systematic review to investigate and quantify the effect of peer‐led interventions that sought to prevent tobacco, alcohol and/or drug use among young people aged 11–21 years.MethodsMedline, EMBASE, PsycINFO, CINAHL, ERIC and the Cochrane Library were searched from inception to July 2015 without language restriction. We included randomized controlled trials only. Screening and data extraction were conducted in duplicate and data from eligible studies were pooled in a random effects meta‐analysis.ResultsWe identified 17 eligible studies, approximately half of which were school‐based studies targeting tobacco use among adolescents. Ten studies targeting tobacco use could be pooled, representing 13 706 young people in 220 schools. Meta‐analysis demonstrated that the odds of smoking were lower among those receiving the peer‐led intervention compared with control [odds ratio (OR) = 0.78, 95% confidence interval (CI) = 0.62–0.99, P = 0.040]. There was evidence of heterogeneity (I 2 = 41%, χ2 15.17, P = 0.086). Pooling of six studies representing 1699 individuals in 66 schools demonstrated that peer‐led interventions were also associated with benefit in relation to alcohol use (OR = 0.80, 95% CI = 0.65–0.99, P = 0.036), while three studies (n = 976 students in 38 schools) suggested an association with lower odds of cannabis use (OR = 0.70, 0.50–0.97, P = 0.034). No studies were found that targeted other illicit drug use.ConclusionsPeer interventions may be effective in preventing tobacco, alcohol and possibly cannabis use among adolescents, although the evidence base is limited overall, and is characterized mainly by small studies of low quality.