Plasma ACTH and serum cortisol levels were measured at 20-min intervals for 24 h in six young women with unipolar endogenous depression and in eight normal women during the early follicular phase of the menstrual cycle. The women with depression had a marked increase (P less than 0.005) in mean ACTH pulse frequency [14.5 +/- 0.6 (+/-SE) pulses/24 h] compared with normal women (9.9 +/- 0.7 pulses/24 h), while mean ACTH pulse amplitude and 24-h transverse mean ACTH levels were similar in the two groups. In contrast, 24-h transverse mean cortisol levels were higher (P less than 0.02) in the depressed women (242 +/- 28 nmol/L) than in the normal women (163 +/- 10 nmol/L). This hypercortisolemia in the depressed women was accompanied by markedly increased (P less than 0.001) episodic cortisol secretion (286 +/- 24 X 10(2) nmol/L X min) compared with that in normal women (155 +/- 17 X 10(2) nmol/L X min), and the secretory episodes were both longer in duration (P less than 0.05) and of higher amplitude (P less than 0.05) in the depressed women. The circadian variations in ACTH and cortisol were maintained in these depressed women, and the times of the circadian nadir, as determined by cosinor analysis, were similar to those in the normal women. However, the mean length of the evening quiescent period of cortisol secretion was far shorter (P less than 0.005) in the depressed women (27 +/- 8 vs. 202 +/- 40 min). Moreover, the postlunch rise in serum cortisol was significantly higher (P less than 0.02) in the depressed women (204 +/- 29 vs. 111 +/- 15 nmol/L). These results provide evidence that the hypercortisolism in depressed women is associated with an increase in ACTH pulse frequency, expanded cortisol secretory episodes, including a greater postlunch rise in cortisol, and a shortened evening quiescent period of cortisol secretion. Our findings provide evidence for centrally mediated activation of the ACTH-cortisol system in women with depression without a phase shift in circadian rhythm.