ORMAL in essentially all people who have no psychiatric or medical illness,‘.’ the dexamethasone suppression test is abnormal in about half of patients with an endogenous depressive illness. That this fraction is not li 10 or 9/10 has made its potential role in the diagnostic process less obvious. On the other hand, since the endogenous depressions which have been studied with this test were so classified with only clinical symptoms and signs, it is easy to wonder if patients with genuine biochemically derived depressions would be nearly unanimous in showing abnormal dexamethasone suppression tests. Under such a hypothesis, essentially all depressed patients with normal suppression of serum cortisol might be thought of as having a severe psychogenic depression which clinically mimics the symptoms and signs by which endogenous depression is diagnosed. It is also conceivable (but not proven) that in some psychogenic depressives abnormal results of dexamethasone suppression tests might develop from somatic stresses of poor nutrition 01 changes in the circadian pattern of activity and sleep, for example. To test the hypothesis that essentially all patients with biochemically derived depressions have abnormal dexamethasone suppression tests, the selection of suitable patients must be rigorous. Childbirth is a somatic experience which generates severe depressive psychoses which would not otherwise 0ccur.j In particular, the frequency of onset of depressive illness during the period between two days and two weeks after parturition is so much larger than the expected incidence among non-postpartum females that essentially all of the depressive illness beginning during this interval can be regarded as depression generated by parturition. In this study it will be regarded that an incapacitating postpartum depression of onset between 2 days and 2 weeks after delivery is a biochemically derived depressive illness. A study of the dexamethasone suppression test among postpartum women would probably be complicated by the effects on cortisol regulation of the many stressful somatic experiences surrounding childbirth. This hormonal test of excess cortisol secretion can be abnormal from surgery or hypoglycemia.‘.’ The stress of labor itself generates high levels of ACTH and cortisol.‘.’ The puerperium is a time of rapid large changes in levels of various hormones. such as progesterone, the estrogens, prolactin, and oxytocin; there are many possibilities of effects of hormones on each other which might interfere with