This paper reviews recent literature which suggests that sleep disturbance in members of the general population, whether or not they have ever had a formal psychiatric disorder, is a risk factor for the onset of a formal psychiatric diagnosis at a later time. Based upon the current literature, the strongest link is between subjective insomnia, lasting at least 2 weeks, and the later onset of depression. Less well-established data suggest that lifetime reports of at least 2 weeks of insomnia, hypersomnia, or both hypersomnia and insomnia, are risk factors for the later development of depression, anxiety disorders or substance abuse. More tentatively, preliminary data suggest that increasing subjective sleep disturbance may signal a relapse in remitted depressed patients. Sleep disturbances are common manifestations of major depressive and anxiety disorders. Therefore, sleep complaints may be among the most robust prodromal symptoms reflecting partial depressive or anxiety disorders, which eventually declare themselves as full-blown clinical episodes.