The effect of exercise on negative affect has been examined in hundreds of studies. However, the effect of exercise on diagnosed clinical depression has received far less attention. Furthermore, poor methodological techniques predominate and results have been conflicting. A meta-analysis was conducted to investigate the effect of exercise on clinical depression and depression resulting from mental illness. The chosen studies examined the effect of a chronic exercise paradigm (independent variable) on depression (dependent variable). Each study’s variables were coded: design, subjects, exercise, and dependent measure characteristics that could moderate the effect of exercise on depression. Moderator variables were analyzed using analysis of variance (ANOVA). Results from 30 studies showed an overall mean effect of −.72. Therefore, individuals who exercised were −.72 of a standard deviation less depressed than individuals who did not exercise. Moderating variables and implications for the prescription of exercise as an effective treatment for depression are discussed.