For any diagnostic system to be clinically useful, and go beyond description, it must provide an understanding that informs about aetiology and/or outcome. DSM-III and DSM-IV have provided reliability; the challenge for DSM-V and DSM-VI will be to provide validity. For DSM-V this will not be achieved. Believers in DSM-III and DSM-IV have impeded progress towards a valid classification system, so DSM-V needs to retain continuity with its predecessors to retain reliability and enhance research, but position itself to inform a valid diagnostic system by DSM-VI. This review examines the features of a diagnostic system and summarizes what is really known about mood disorders. The review also questions whether what are called mood disorders are primarily disorders of mood. Finally, it provides suggestions for DSM-VI.