Neuroendocrine abnormalities in depression have been regarded, by many authors, as relatively specific markers of nosological subtypes of the disorder, e.g. primary vs. secondary, endogenous vs. non-endogenous or unipolar vs. bipolar depression. They should reflect the same changes in central neurotransmitters (e.g. noradrenergic insufficiency and/or cholinergic hyperactivity) that were hypothesized as the cause of clinical symptoms. This view is challenged on the basis of our own neuroendocrine investigations in 317 psychiatric patients and 103 normal controls. According to these studies the abnormalities are nosologically rather unspecific. They are induced by a large variety of factors, e.g. emotional stress associated with the clinical symptomatology, weight loss due to malnutrition as a consequence of reduced appetite, medication and drug withdrawal. Stress-induced hypercortisolism appears to be the most common abnormality that may trigger other neuroendocrine dysfunctions, such as a blunted TSH response to TRH. Differences in neuroendocrine abnormalities of depressives are probably due to variations in the manifold factors influencing the hormonal axes involved, to temporal changes in hormonal patterns (e.g. one abnormality triggering another) and to individual differences in the basic activity and the responsiveness of the various axes.