The two main classifications of mood disorders currently used are the American system, DSM, the fourth edition of which came out in 1994, and the International Classification of Diseases, the tenth edition of which was published in 1993. These classifications are based on the following broad principles: simple description with no hypothetical aetio-pathogenic discussion, distinction between the various clinical forms in accordance with variety and intensity of symptoms, co-existence of other more or less serious somatic disorders, individual identification of bipolar forms, etc. The key changes made in the classification of mood disorders between the revised version of DSMIII published in 1987 and DSM-IV published in 1994 suggest that a greater degree of concordance in diagnosis between different practitioners may be expected in future. In addition, certain changes such as the therapeutic decision-trees will most likely have an impact on practice. In this way, further details are added to the clinical criteria of duration or description of the clinical features seen, thereby allowing a clearer distinction to be made between normality and the onset of the pathology in question. Regarding bipolar disorders, the general organisation of the classification and the terminology used have been extensively revised, with the distinction between types I and II becoming official. Furthermore, differentiation between certain aspects of depression emphasises the frequency of certain clinical particularities, either in the onset or during the course of these disorders. This is true of melancholic, catatonic and atypical features specifiers (i.e. with mood reactivity and interpersonal rejection sensitivity), post-partum onset specifier, seasonal pattern specifier and rapid-cycling specifier. Finally, a certain number of specifications are proposed allowing the postulation of notions of complete or partial cure between episodes in case of recurrence.