BACKGROUND: We evaluated the occurrence of postnatal depression in general and during different seasons as part of a larger longitudinal mother-child follow-up study. METHOD: One hundred and eighty-five mothers, from the maternity wards of University Hospital of Oulu, Finland, completed a self-rating depression scale, the Edinburgh Postnatal Depression Scale (EPDS) twice: first at hospital 2-7 days after delivery and the second time at home 4 months after the delivery. Different psychosocial variables were mapped out to avoid any confounding factors. The year was divided in two separate ways: first, three different time periods were selected by the amount of sunlight: dark (October-January), intermediate (February, March, August, September) and light (April-July), and second, the year was divided by seasons. The results were analysed by the chi(2)-test for multinomials. RESULTS: Sixteen percent (16.2) of mothers were scored as being depressed using 13 as a cut-off point immediately after the infant was born. Thirteen percent (13.0) were depressed measured 4 months postpartum. There was more mild depression in the autumn (ratio observed/expected 1,62; 95% confidence interval 1.05-2.19) immediately after delivery, using 10 as a cut-off, and less depression in the spring (0.27; 0.00-0.62) measured at home later, using 13 as a cut-off. When using classification by the amount of light there was more depression during the dark time (1.58; 1.05-2.11) immediately postpartum. LIMITATIONS: The group sizes and the amount of sample sizes collected within each month are quite small. CONCLUSIONS: It should be borne in mind that seasonal changes and alterations in the amount of light might influence the occurrence of postnatal depression.