BackgroundPostpartum depression (PPD) is an important public health problem worldwide. China is planning to launch PPD screening in community settings, but there are questions on the community prevalence of PPD and validated screening tools. MethodsWe sought to recruit all eligible new mothers during postnatal home visits in two districts of Changsha, China, and after informed consent, screened them for PPD using three self-administered questionnaires—the Edinburgh Postpartum Depression Scale (EPDS), the Patient Health Questionnaire (PHQ-9), and Whooley Questions for Depression Screening. Video structured diagnostic interviews were performed online according to Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV) by qualified specialists who were blinded to screening results. Optimal screening was determined based on the acceptability of scales and diagnostic accuracy metrics including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). ResultsOut of 3004 eligible women, 2730 (90.9 %) completed the screening questionnaires. Among those screened, the video structured diagnostic interview was administered to 1862 (68.2 %) and 62 (3.3 %) were diagnosed with a current depressive condition. The optimal screening approach involved combining Whooley Questions (at least one “yes”) with EPDS (cutoff >10) in series, with sensitivity of 0.76 (95 % CI 0.63 to 0.85), specificity of 0.93 (0.92 to 0.94), PPV of 0.28 (0.21 to 0.36) and NPV of 0.99 (0.98 to 1.00). LimitationsDue to the regional sample and exclusion of mothers with telephone contact rather than home visits, our findings may not be fully generalizable to the entire population. ConclusionsThe prevalence of PPD among women in this sample was substantially lower than those reported in previous studies in China, the majority of which used screen positivity in measuring prevalence. Combining Whooley Questions with EPDS in series is the most optimal screening approach in this population, though this would still result in a high number of false positives at current prevalence.