Current theories of the etiology of postpartum depression provide three models for understanding women's affective functioning during the postpartum period: a biological model, which emphasizes the role of hormonal changes; a psychological model, which gives primacy to women's enduring personal characteristics; and a stress model, which construes postpartum depression as the outcome of the stress of childbirth in interaction with other factors. In this article I criticize these theories, based on anthropological and psychological research in a rural Kipsigis community of Kenya. Data on the cultural structuring of pregnancy, childbirth, and the postpartum period among the Kipsigis constitute the principal source for this critique. In addition, the affective functioning of a sample of ten women in the community during pregnancy and the postpartum period was studied using reports of memories and dreams as measures of affective valence and thematic concern. The women's reports are compared to each other at three different time points (during the last half of pregnancy, at two to three weeks postpartum, and at two to three months postpartum), and to responses of a matched comparison group. At both the cultural and individual levels, there seems to be no evidence of postpartum depression in this setting, and it is suggested that culture is a powerful mediating factor between the physiological processes related to childbirth and their psychological outcomes.