This study aimed to evaluate how precise the Edinburgh Depression Scale (EDS) is in screening for major depressive disorder (MDD) during different periods of pregnancy. A random sample of 230 pregnant women was interviewed in the first, second, and third trimesters of pregnancy using the EDS and not-patient version of the Structured Clinical Interview for DSM-III-R (SCID-NP). We evaluated test-retest reliability of the EDS; area under the ROC curve (AUC), sensitivity, specificity, and positive predictive value (PPV) of the EDS against the SCID-NP diagnoses in the first, second, and third trimesters of pregnancy. Test-retest reliability of the EDS was 0.81 (p < 0.001). An optimal cutoff score of the EDS for screening current SCID-NP diagnosis of MDD was 12 and higher in the first trimester of pregnancy (AUC 0.94, sensitivity 92%, specificity 95%, and PPV 52%) and 11 and higher in the second and third trimesters of pregnancy (AUC 0.96 and 0.90, respectively; sensitivity 100% and 88%, respectively; specificity 92% and 92%, respectively; PPV 25% and 29%, respectively). The EPDS is a reliable instrument for repeated evaluations of depressive symptoms during pregnancy. It has a good sensitivity and specificity for detecting antenatal MDD with optimal cutoff of 11/12 or higher.