•The pathogenesis and progression of depression may depend on different lifestyle determinants related to depression.•The lifestyle medicine approach may be of great potential to be implemented as an entry-level community intervention to prevent depression and an adjunctive intervention in managing major depression.•The number of lifestyle factors adopted was a significant moderator for the effect of multi-component lifestyle medicine interventions on depression.•The clinical effect of multi-component lifestyle interventions tend to be stronger in major depression.•Smartphone-delivered lifestyle medicine interventions are needed to increase accessibility and support intervention compliance. The treatment effect of multi-component LM interventions on depressive symptoms has not yet been examined. We systematically searched six databases from inception to February 2020 to identify randomized controlled trials (RCTs) involving any multi-component LM interventions (physical activity, nutritional advice, sleep management, and/or stress management) on depressive symptoms relative to care as usual (CAU), waitlist (WL), no intervention (NI), or attention control (AC) comparisons. Fifty studies with 8,479 participants were included. Multi-component LM interventions reduced depressive symptoms significantly relative to the CAU (p >.001; d = 0.20) and WL/NI (p > .01; d = 0.22) comparisons at immediate posttreatment. However, no significant difference was found when compared with AC. The intervention effects were maintained in the short-term (1- to 3-month follow-up) relative to the CAU comparison (p > .05; d = 0.25), but not in the medium- and long-term. The moderator analyses examining the effect of multi-component LM interventions compared with CAU suggested that the number of lifestyle factors adopted was a significant moderator. Although disease type was not a significant moderator, there was a tendency that the clinical effect of multi-component LM interventions was stronger (d = 0.45) in those diagnosed with major depression. No publication bias was detected. Low number of RCTs available in some subgroup analyses prevented from finding meaningful effects. Results may not be extended to major depression, because data on secondary depression were captured. Multi-component LM interventions appeared to be effective in mitigating depressive symptoms; however, the magnitude of the clinical effect was small. Future research is needed to assess more comprehensive and individualized LM interventions which have a greater emphasis on motivational and compliance aspects and focus solely on individuals with depression.