Objective We examined the effects of exercise on depression in older adults living in nursing homes and explored the moderator effects of participants, methods, and intervention characteristics. Methods We searched 8 databases from inception to January 2020 without date restrictions. We retrieved primary studies measuring exercise with 60-year-olds with depression that were written in English. Two researchers independently coded each primary study and compared codes for discrepancies. They consulted a third researcher to come to consensus. We used random-effects model to compute effect sizes using Hedges' g, a forest plot, and Q and I-2 statistics as measures of heterogeneity. We also examined moderator analyses. Results Twenty-one studies included 2,594 participants (81.8 +/- 4.8 years old). Overall, exercise showed a small improvement in depression compared to controls (ES=.25; 95%CI 0.11, 0.38; p=.000). Interestingly, interventionists by physiotherapists had a smaller effect on depressive symptoms (.04) than other health interventionists (.37). As people aged, exercise was less effective in reducing depression (slope=-.03, Q(model)=5.03, p=.025). Researchers who included a higher percentage of women in exercise showed less improvement of depression (slope=-.01, Q(model)=8.60, p=.003). Also, when researchers used depression scales with higher reliability values, they measured higher levels of depression (slope = 5.48, Q(model)=5.60, p=.018). No other quality indicators moderated the effects of exercise on depression. Conclusion Exercise significantly improved depressive symptoms among older adults in nursing homes. Exercise might be used as adjunct/alternative complementary treatment for improving depressive symptoms in older adults in nursing homes.