Mental health difficulties are prevalent in autistic people with [approximately]14%-50% having experienced depression and [approximately]40%-80% having experienced anxiety disorders. Identifying interventions that improve autistic people's mental health is a top priority. However, at present, there is no high-quality network meta-analysis of benefits and harms of different interventions. We conducted a systematic review and network meta-analysis of randomised controlled trials, searching MEDLINE, EMBASE, other databases, and trial registers until 17 October 2020. We included randomised controlled trials reporting anxiety or depression in a suitable format. We calculated effect estimates and 95% credible intervals using Bayesian network meta-analysis. Our search identified 13,794 reports, of which 71 randomised controlled trials (3630 participants) were eligible for inclusion. All trials had high risk of bias. The follow-up period ranged from 1 to 24 months. Evidence indicates uncertainty about the effects of different interventions, with more high-quality evidence needed. Available evidence suggests that some forms of cognitive behavioural therapy may decrease anxiety and depression scores in autistic children and adults; mindfulness therapy may decrease anxiety and depression scores in autistic adults with previous mental health conditions; and behavioural interventions may provide some benefit for depression in autistic children. We recommend that autistic people are given access to mental health interventions available to non-autistic people, following principles of person-centred care.