variety of stimuli, including exercise, but the mechanisms by which running affects neurogenesis are not yet fully understood. Because beta-endorphin, which is released in response to exercise, increases cell proliferation in vitro, we hypothesized that it could exert a similar effect in vivo and mediate the stimulatory effects of running on neurogenesis. We thus analyzed the effects of voluntary wheel-running on adult neurogenesis (proliferation, differentiation, survival/death) in wild-type and beta-endorphin-deficient mice. In wild-type mice, exercise promoted cell proliferation evaluated by sacrificing animals 24 h after the last 5-bromo-2'-deoxyuridine (BrdU) pulse and by using endogenous cell cycle markers (Ki67 and pH(3)). This was accompanied by an increased survival of 4-wk-old BrdU-labeled cells, leading to a net increase of neurogenesis. beta-Endorphin deficiency had no effect in sedentary mice, but it completely blocked the running-induced increase in cell proliferation; this blockade was accompanied by an increased survival of 4-wk-old cells and a decreased cell death. Altogether, adult neurogenesis was increased in response to exercise in knockout mice. We conclude that beta-endorphin released during running is a key factor for exercise-induced cell proliferation and that a homeostatic balance may regulate the final number of new neurons.