For some 20 years, numerous research teams have sort to identify specific biological markers for depression. There have been three main stages in this search, each of which was characterized by progress in the technologies then available. The first stage involved evaluation of the activity of known neurotransmitters, involving assay of concentrations of precursors or catabolites, as well as measurement of the activity of enzymes regulating synthesis and catabolism, rate of up-take and storage. The second phase involved use of specific radioactive ligands which bind to neuronal membranes and receptors. This allowed in vivo exploration of the central nervous system in man. Finally, the third stage has involved a functional approach, both cellular and systemic, based upon the dynamic measurement of biological response to stimuli of central origin. While the first two approaches are of course a rather indirect reflection of central nervous system function, the third allows assessment interactions between the various systems. The countless laboratory studies carried out in depression have not provided useful etiologic, diagnostic or therapeutic results, and currently, far more attention is given to the interactions between the various neurotransmitter systems than to each system taken individually. Numerous studies have been carried out to assess interactions between the neurological, immunological and endocrine systems, and abnormal immunological parameters have been reported. Similarly, the endocrine disorders seen in depression could be attributable to a disorganization of the hippocampal-hypothalamic-adrenalin axis due to accumulation of stress. The vulnerability could also be due to preclinical sensitization, as described in the "kindling" model. Post gave particular importance to the role of biochemical modifications induced by repetitive stress. Long-term modifications could involve changes in transcription factors involved in protein synthesis, leading to long-term changes in peptides which, in turn, could lead to neurobiological modifications in the limbic system, a sort of "depression memory" rendering patients far more vulnerable to subsequent episodes. Finally, we would note that the identification of a gene for susceptibility to disorders of mood in some families is only a first step towards the characterization of the protein produced by the gene; this protein should in turn either modulate or produce a signal or a message of some sort. Furthermore, the genetic component itself could act via environmental or relational factors involved in the genesis of the depression. Finally, it is very important to bear in mind the great plasticity of neuronal systems, which are continuously subject to adaptation and modulation phenomena.