Abstract Background This study aimed to assess the impacts of closed-off measures with different strictness levels (lockdown, partial lockdown and non-lockdown) and geographic proximity to patients with coronavirus disease 2019 (COVID-19) on prenatal depression during an epidemic rebound of COVID-19. Methods This was a cross-sectional web-based survey including 880 pregnant women. Depressive symptoms were measured by Self-Rating Depression Scale (SDS) and geographic proximity was calculated using Geographic Information Systems. Linear and logistic regression were used to assess the associations of closed-off measures and geographic proximity with SDS scores and depressive symptoms. Restricted cubic splines were used to model non-linear associations between geographic proximity and depression symptoms. Results Compared with those living in non-lockdown areas, women in lockdown areas had higher SDS scores (adjusted β: 3.51, 95% CI: 1.80, 5.21) and greater risk of depressive symptoms (adjusted OR: 4.00, 95% CI: 2.18, 7.35), but evidence for partial lockdown was not obvious. A progressive increase in the risk of depressive symptoms was found with decreasing distance to COVID-19 patients when geographic proximity was