Introduction According to many studies, anxiety in the perinatal period is widespread and has many detrimental effects. Thus, screening measures should not be limited to assessing depression symptoms. The widely used Edinburgh Postnatal Depression Scale (EPDS) might assess depression but also anxiety symptoms. This study explores whether an anxiety dimension (EPDS-3A) was found and valid in French women during pregnancy and the postpartum period. Methods French women were followed-up at late pregnancy and 2 and 4 months postpartum (N = 144, 138 and 129). They completed the EPDS and the Hospital Anxiety and Depression Scale (HADS-A). Exploratory factor analyses were performed. Then to test its validity, the EPDS-3A was correlated with anxiety (HADS-A) and depression (EPDS-D) scores. Finally, prevalence estimates were computed according to recommended cut off. Results The anxiety dimension assessed through the EPDS-3A was observed during the postpartum period but not during pregnancy. A two-factor structure (depression and anxiety) increases the variance explained at 2 and 4 months postpartum (respectively 6 and 12%). The EPDS-3A shows good internal consistency (≥ .70) and was more strongly associated with anxiety scores (HADS-A) (.48–.57) than with depression scores (EPDS-D) (.30–.39). Nearly 28% of mothers had scores that exceeded the EPDS-3A cut off (≥ 4) but not the full EPDS cut off (≥ 13 or more). Discussion The EPDS contains an anxiety component (EPDS-3A) that can be found in French women during the postnatal period but not during pregnancy. It shows signs of internal consistency and validity. The EPDS-3A could be considered when screening for postpartum anxiety.< Réduire