Late-life depression is a growing public health concern. Exercise may be of added value but the literature remains equivocal. We conducted a systematic overview of meta-analyses and an exploratory pooled analysis of previous meta-analyses to determine the effect of exercise on depression in older adults. Two independent researchers searched Pubmed, CINAHL, Cochrane Plus, PsycArticles, and PsycInfo for meta analyses on exercise in late-life depression. Methodological quality was assessed using the Assessment of Multiple Systematic Reviews (AMSTAR) Instrument. We pooled effect sizes from previous meta-analyses of randomized controlled trials to determine the effect of exercise on depression in older adults. The systematic review yielded 3 meta-analyses. In total, 16 unique cohorts of 1487 participants were included. The quality of the three included meta-analyses was considered as "moderate" according to AMSTAR scores. No serious adverse events were reported. Compared to controls (n=583), those exercising (n=541) significantly reduced depressive symptoms. Our umbrella review indicates that exercise is safe and efficacious in reducing depressive symptoms in older people. Since exercise has many other known health benefits, it should be considered as a core intervention in the multidisciplinary treatment of older adults experiencing depression. (C) 2016 Elsevier Ireland Ltd. All rights reserved.