A sample of over 6000 women was recruited in the last trimester of pregnancy and administered a 40-item self-report questionnaire designed to detect the presence of factors that were likely to increase the risk of postpartum depression. The mental state of almost 5000 of these women was determined at around 6 to 8 weeks postpartum. By conducting a series of logistic regressions on two-thirds of this sample the original set of variables was reduced to a predictive index of 17 items with weighted scores calculated for each. This index was then applied to the remaining one-third of the sample as a validating procedure and specificity and sensitivity calculated. The index offers a system for the prediction of postpartum depression that could be of use in both research and clinical practice.