BACKGROUND: A correlation between gonadal steroids and depressive symptoms during the perinatal period has long been suggested; however, the underlying mechanism for this relationship remains unclear. METHODS: This study was designed to examine the correlation between gonadal steroid concentrations of umbilical cord blood and postpartum depressive symptoms as well as longitudinal alterations in maternal plasma gonadal steroid concentrations among 204 perinatal women. The levels of postpartum depressive state at 1 month postpartum were evaluated using the Edinburgh Postnatal Depression Scale. RESULTS: Umbilical progesterone, estradiol, and testosterone levels were significantly higher in infants delivered by depressed mothers (870.7 ± 281.7 ng/ml, 8607.7 ± 4354.6 pg/ml, and 2.5 ± 0.9 ng/ml, respectively) than those delivered by nondepressed mothers (741.3 ± 324.0 ng/ml, 5221.9 ± 3416.3 pg/ml, and 2.1 ± 0.6 ng/ml, p < .01, p < .05, and p < .05, respectively). Postpartum plasma progesterone levels of depressed mothers (3.5 ± 3.1 ng/ml) measured in the early postpartum period were significantly lower than those of nondepressed mothers (9.1 ± 9.7 ng/ml, p < .01). The decrease in progesterone from mid-pregnancy to the early postpartum period was significantly higher in depressed mothers than in nondepressed mothers. Subgroup analyses specific to primiparas or multiparas indicated that a significant drop of progesterone was seen only in primiparas. CONCLUSION: The current study suggests that the delivery of a placenta/fetus with high gonadal steroid production may cause a wider range of fluctuations in maternal plasma gonadal steroid concentrations, which may be concurrent with postpartum depressive symptoms.