To evaluate the reliability of the endogenous concept of depressive illness according to the Newcastle Endogenous Depression Diagnostic Index (NEDDI), 155 major depressive inpatients with NEDDI scores > or = 6 (endogenous) were matched for gender and age (+/- 5 years) to 155 major depressive inpatients with NEDDI scores < 6 (nonendogenous). When sleep polygraphic variables, neuroendocrine parameters (dexamethasone suppression and thyrotropin-releasing hormone tests), and various clinical variables (unipolar/bipolar status, psychotic/nonpsychotic subtype, and severity of the depressive episode) were examined, statistically significant differences between endogenous and nonendogenous patients emerged for three variables: the thyroid-stimulating hormone response to the thyrotropin-releasing hormone test, the dexamethasone suppression test response at 16:00 h, and the percentage of time awake during the night. However, when the effects of age and severity of depression were controlled, those differences disappeared.