In this article, we systematically reviewed the results of application of biological markers of depression to children and adolescents. Concerning sleep EEG, only three studies on a total of twelve among 267 depressed children and adolescents aged 6 to 19 years found the typical sleep abnormalities described in depressed adults (eg, shortened REM latency and decreased sleep efficiency). Most authors insisted on the age-related sleep changes as a major confounding factor. Two studies of the effect of antidepressant therapy on sleep showed a decrease in sleep efficiency but a discrepancy in the evolution of REM latency. Concerning the dexamethasone suppression test, twenty studies including 374 depressed children and adolescents (3-20 years) and 533 psychiatric controls yielded an overall sensitivity of 57% and an overall specificity of 78%. These results may be considered as interesting, despite the lack of agreement among authors on various methodological parameters (dose of dexamethasone, times of blood sampling, method of cortisol assay ...) and the composition of control groups which often comprise subjects presenting disorders very close to major depression (dysthymia, minor depression ...). Among the other tests, the TRH test, used in two studies, showed limited interest. In contrast, the study of growth hormone secretion, performed in one centre, could present diagnostical usefulness. In conclusion, biological markers of depression in children and adolescents should still be considered as research tools and be part of a multidisciplinary approach to depressive illness.