Increased plasma cortisol in patients with major depression is a well documented finding, although it is present in only 25-30% of subjects with major depression. However, ACTH and cortisol are secreted in a pulsatile manner, so it is unclear if increased ACTH secretion occurs in depression and if there are changes in the pulsatile components of ACTH secretion. Ten-minute sampling for ACTH and cortisol was performed for 24 hr in 25 premenopausal depressed women, whose age and menstrual cycle day matched control women. As a group, the depressed women demonstrated a trend to increase cortisol secretion (p = 0.089). There was no difference in mean cortisol between the patient group as a whole (8.36 +/- 2.9 microg/dl) and those patients meeting criteria for atypical depression (8.38 +/- 1.9 microg/dl), but patients meeting criteria for endogenous showed increased cortisol (12.17 +/- 4 microg/dl) Mean ACTH was not significantly different between patients and controls. Pulse analyses revealed similar number of secretory events and similar amplitudes for cortisol secretory bursts in patients and controls. The baseline component area under the curve of cortisol secretion was increased at a trend level (p =.064) in depressed patients, and the baseline AUC for ACTH was significantly increased in depressed patients (p =.045). No differences were found in pulsatile components of ACTH secretion between patients and matched controls. Harmonic analyses indicated no significant differences between patients and controls on any detected rhythm for ACTH or cortisol. These data suggest that the pulsatile and circadian components of the HPA axis are normal in premenopausal depressed women and that only 24% of depressed women demonstrate hypercortisolemia.