This study of 51 prepubertal psychiatric inpatients evaluates plasma cortisol measurements at 8 AM, 4 PM, and 11 PM before and after dexamethasone was administered in counterbalanced order at doses of 0.5 mg and 1.0 mg. Approximately 76.5% of the children had an affective disorder. Major depressive disorder was associated with higher plasma cortisol levels than other disorders. Pre- and postdexamethasone plasma cortisol levels using 0.5 mg dexamethasone exhibited a circadian variation. The optimal criterion for cortisol nonsuppression was 5 micrograms/dl measured at 8 AM after administration of 0.5 mg dexamethasone.