Voluntary exercise has been reported to reduce depression- and anxiety-like behaviors in an animal model of depressive disorder (DD). However, the degree of appropriate voluntary exercise to improve depression-like behavior remains unknown. In the present study, we examined the effects of two different terms of voluntary exercise, short (3 days)- and long (14 days)-term freewheel running, on depression-like behavior and new cell proliferation in the dentate gyrus (DG) in a DD mouse model generated by exposure to chronic mild stress (CMS), and then compared the antidepressant effects with those of ketamine. Short-term freewheel running showed an antidepressant effect in CMS mice in the forced-swim test (FST), which was comparable to that of ketamine, but did not affect new cell proliferation in the DG. In contrast, long-term freewheel running increased new cell proliferation in the DG, but did not improve depressive behavior in the FST. These results suggest that voluntary exercise, unless taken to the point of exhaustion, could beneficially contribute to the recovery of CMS-induced depression.