• To ascertain the extent of dysregulation of melatonin secretion in endogenous depression, we measured nocturnal and diurnal serum melatonin concentrations in 38 depressed patients (23 women and 15 men) who had primary, definite endogenous depression according to the Research Diagnostic Criteria and in 38 individually matched normal control subjects. Previous reports have suggested that such patients may have reduced nocturnal melatonin secretion, often in conjunction with increased hypothalamicpituitary-adrenal cortical axis activity. This has been considered as a possible reflection of reduced noradrenergic activity in depression. Compared with their matched controls, the depressed patients showed a trend toward a significantly elevated average nocturnal melatonin concentration that was accounted for primarily by the 14 premenopausal women—the postmenopausal female and male depressive patients did not differ significantly from their respective controls. The average diurnal melatonin concentration also showed a trend toward being higher in both the female and male depressed patients. The melatonin measures were not consistently related to any of the previously reported hypothalamic-pituitary-adrenal cortical axis measures in these subjects. Our findings thus failed to confirm a "low melatonin syndrome" or an inverse relationship between nocturnal melatonin and nocturnal cortisol concentrations in depression. This discrepancy may be related to methodologic differences among studies; our data are in accord with those findings of the one other reported study in which normal controls were individually matched to patients on variables that were known to influence melatonin secretion. Most of the studies, including ours, have been cross-sectional. However, the few longitudinal studies to date have not found consistent differences in melatonin profiles in the same patients when they were depressed and when they were euthymic or manic.