The current study examined the role of cognitive factors in the development and maintenance of depressive symptoms from pregnancy into the postpartum period. One hundred and one women were assessed for levels of rumination (brooding and reflection), negative inferential styles, and depressive symptoms in their third trimester of pregnancy and depressive symptom levels again at four and eight weeks postpartum. We found that, although none of the three cognitive variables predicted women's initial depressive reactions following childbirth (from pregnancy to one month postpartum), brooding rumination and negative inferential styles predicted longer-term depressive symptom changes (from pregnancy to two months postpartum). However, the predictive validity of women's negative inferential styles was limited to women already exhibiting relatively high depressive symptom levels during pregnancy, suggesting that it was more strongly related to the maintenance of depressive symptoms into the postpartum period rather than increases in depressive symptoms following childbirth. Modifying cognitive risk factors, therefore, may be an important focus of intervention for depression during pregnancy.