Objective: This study aimed to elucidate whether sufficient sleep in each stage of pregnancy impacted on reducing PPD, and whether changes in sleep pattern during the prenatal period was associated with PPD, using a longitudinal and large cohort. Methods: The study participants were recruited at two special hospitals from Mar 2013 to Nov 2017, and included Korean pregnant women around 12 gestational weeks (GW); women with pregnancy of triplets or quadruplets were excluded. They completed a sleep questionnaire regarding whether or not they had sufficient sleep at 12 GW, 24 GW, and 36 GW. Participants were divided into Group0 (sustained sufficient sleep), Group1 (change from sufficient to insufficient sleep), Group2 (change from insufficient to sufficient sleep), and Group3 (sustained insufficient sleep). Depressive symptoms were assessed by the Edinburgh Postnatal Depression Scale (EPDS) at postpartum 4 weeks and cutoff score for PPD was ≥10. Results: Out of 2512 participants, 410 (16.3%) women were diagnosed with PPD. Only sufficient sleep at 36 GW was significantly associated with lowering PPD, after adjusting for confounding factors in multivariate regression analysis (odds ratio, 0.560; 95% confidence interval, 0.440-0.714; P < .001). When the sleep change between three time points (pre-pregnancy, 12 GW, or 24 GW) and 36 GW was assessed, both Group1 (worsening sleep) and Group3 (sustained insufficient sleep) were significantly associated with developing PPD at all starting time points in the multivariate analysis, while there was no significant association between Group2 (improving sleep) and PPD. Conclusion: Sufficient sleep in late pregnancy (at 36 GW) had the impact of lowering PPD. Whether or not participants had sufficient sleep at the pre-pregnancy or prenatal periods until 36 GW, having insufficient sleep at 36 GW was highly associated with developing PPD.