OBJECTIVE: To develop and evaluate a questionnaire used antepartum to screen for postpartum depression. STUDY DESIGN: Demographic and clinical data, based on previously identified variables, were obtained from 106 second-trimester gravidas (sample I) by interview, self-administered questionnaire and medical record review. The Beck Depression Inventory (BDI) was administered at 1, 6 and 12 weeks postpartum (PP). Statistical analysis, including stepwise linear regression with maximum r-squared improvement, identified a subset of the 24 most predictive variables. This antepartum questionnaire (APQ) was validated retrospectively in the original sample and prospectively in a second group of 99 women (sample II). RESULTS: In both sample populations the APQ had acceptable sensitivity (80-82%) and specificity (78-82%). The incidence of postpartum depressive symptoms (PPDS) rose from 10% to 17% by six weeks without an appreciable decline at 12 weeks (15%). The percentage of women showing more than mild depressive symptoms increased with PP time from 30% at 1 week to 47% at 12. CONCLUSION: The APQ is now available for screening and evaluating early therapeutic intervention in PPDS.