OBJECTIVE:To compare the effectiveness of non-pharmacological interventions on depressive symptoms in people after stroke. DATA SOURCES:A literature search was performed through databases from January 2000 to August 2018: MEDLINE; CINAHL Plus; Scopus; Academic Search Complete; Cochrane Central Register of Controlled Trials; Scopus; and Library, Information Science and Technology Abstracts. Search terms included depression, stroke, non-pharmacologic, and intervention. STUDY SELECTION:We included randomized controlled trials comparing non-pharmacological interventions to controls for depressive symptoms in people after stroke. Of 1703 identified articles, 22 trials were included in narrative synthesis, of which 13 were eligible for meta-analysis. DATA EXTRACTION:Two reviewers extracted characteristics of participants, interventions, and results from all included trials. DATA SYNTHESIS:Thirteen interventions were categorized into four types: complementary and alternative therapy (five trials, n=228), exercise (four trials, n=263), psychosocial therapy (two trials, n=216), and multifactorial therapy (two trials, n=358). Overall beneficial effects of non-pharmacological interventions on depressive symptoms were found both post-intervention (effect size [ES] = -0.24, 95% confidence Interval [CI]: -0.37 to -0.11, p < 0.05) and at follow-up (ES = -0.22, CI: -0.36 to -0.07, p< 0.05). We found individual beneficial effects for complementary and alternative therapy (ES = -0.29, CI: -0.55 to -0.02, p < 0.05) and psychosocial therapy (ES = - 0.33, CI: -0.60 to -0.06, p < 0.05) post-intervention. CONCLUSIONS:Complementary and alternative therapy and psychosocial therapy appear to be promising strategies for improving post-stroke depression. Future studies target a personalized approach for people with specific conditions such as cognitive impairment.