Women with a history of perinatal and nonperinatal major depression are more likely to relapse during pregnancy than postpartum, a finding that points to the need for closely monitoring these women for depression during pregnancy. In addition, these findings of differences in risk factors and clinical features suggest that postpartum-onset major depression may have a pathophysiology distinct from major depression that begins during pregnancy. Time of onset of perinatal depression should be considered in the design of genetic and treatment studies.