Objective: To identify studies of non-pharmacological interventions provided following recovery from depression, and to evaluate their efficacy in preventing further episodes. Method: We identified relevant randomised controlled trials from searching MEDLINE, Embase, PsycINFO, CENTRAL, and ProQuest, searching reference and citation lists, and contacting study authors. We conducted a meta-analysis of relapse outcomes. Results: There were 29 eligible trials. 27 two-way comparisons including 2742 participants were included in the primary analysis. At 12months cognitive-behavioural therapy (CBT), mindfulness-based cognitive therapy (MBCT), and interpersonal psychotherapy (IPT) were associated with a 22% reduction in relapse compared with controls (95% CI 15% to 29%). The effect was maintained at 24months for CBT, but not for IPT despite ongoing sessions. There were no 24-month MBCT data. A key area of heterogeneity differentiating these groups was prior acute treatment. Other psychological therapies and service-level programmes varied in efficacy. Conclusion and implications: Psychological interventions may prolong the recovery a person has achieved through use of medication or acute psychological therapy. Although there was evidence that MBCT is effective, it was largely tested following medication, so its efficacy following psychological interventions is less clear. IPT was only tested following acute IPT. Further exploration of sequencing of interventions is needed