BACKGROUND: Running exercise effectively ameliorates depressive symptoms in humans and depression-like behaviors in animals, but the underlying mechanisms remain unclear. Microglia-mediated neuroinflammation plays a major role in the development of depression. The medial prefrontal cortex (mPFC) is a key brain region involved in depression and is sensitive to physical activity. Whether the antidepressant effect of running exercise involves changes in mPFC microglia is not understood. METHODS: The animals were subjected to chronic unpredictable stress (CUS) intervention followed by treadmill running. The sucrose preference test and elevated plus maze test or tail suspension test were used for behavioral assessment of the animals. The number of microglia in the mPFC was quantified by immunohistochemistry and stereology. The density and morphology of microglia were analyzed via immunofluorescence staining combined with three-dimensional laser scanning techniques. The mRNA expressions of inflammatory cytokines in the mPFC were examined via quantitative real-time PCR. RESULTS: Running exercise effectively alleviated depressive-like behaviors in depression model animals. Running exercise reversed the increase in the number of microglia and the density of activated microglia in the mPFC of CUS animals. Running exercise effectively reversed the changes in microglia (reduced cell body area, total branch length and branch complexity) in the mPFC of CUS animals. Furthermore, running exercise regulated the gene expressions of pro-/antiinflammatory cytokines in the mPFC of CUS animals. CONCLUSIONS: Our results suggested that the antidepressant effects of running exercise may involve decreasing the number of activated microglia, reversing morphological changes in microglia in the mPFC, and reducing inflammatory responses.