PURPOSE: This study aimed to determine the effects of puerperium maternal anxiety, quality of marital relationship, and postpartum blues on late postpartum depression. METHODS: Participants were 130 postpartum women who delivered healthy babies, and answered self-administered questionnaires, such as Edinburgh Postpartum Depression Scale (EPDS-K), Spielberger's State Anxiety Inventory, and Quality of Marital Relationships during the 1st week of delivery and 4–6 weeks postpartum. Data were analyzed using t–test, ANOVA, Spearman's rho correlation coefficient, and hierarchical stepwise multiple regression analysis. RESULTS: The point prevalence of postpartum depression (EPDS-K ≥10, cut-off score) was 23.1% at early postpartum and 16.9% at late postpartum. Predictors of late postpartum depression were early postpartum depression (β=0.86), the difference between early and late postpartum anxiety (β=0.19), and the difference between early and late Quality of Marital Relationships (β=−0.13) (F=70.52, p<0.001). The total vatiance explained was 76.5%. CONCLUSION: It is recommended that a maternal management system should be established at maternity hospitals to screen new mothers' depression and anxiety risks during the early postpartum stage.