To the Editor: In an editorial in the July 2010 issue of the Journal, Gordon Parker, M.D. (1), and a distinguished group of 16 coauthors, including one of the Deputy Editors of this journal, made an argument for melancholia being classified as a distinct mood disorder. The arguments that 1) melancholia features a cluster of symptoms with greater consistency than the broad heterogeneity of the disorders and conditions included in major depression and bipolar disorder and 2) the melancholia diagnosis has superior predictability for prognosis and treatment were not well supported by the evidence. For example, the statement that melancholic patients rarely respond to placebos, psychotherapies, or social interventions could equally well apply to severe major depression and psychotic major depression, both of which are major depression subtypes, as is melancholia, allowed as "specifiers" in the DSM-IV classification of major depression. Relevant literature reveals that hypercortisolemia is not specific to the melancholia diagnosis (2, 3). More importantly, in many studies the melancholia diagnosis lacks predictive value for treatment selection, including response to antidepressant medications or differential predictive value for response across classes of antidepressants (4, 5). Because each of the specifiers designated in DSM-IV shares characteristics with the larger domain of major depression and yet each has its own distinctive qualities (for melancholia it is a characteristic cluster of symptoms), it makes sense to retain the current system in DSM-5.