Although poststroke depression is unlikely to represent a single disorder and numerous etiologies for different kinds of poststroke depression will likely emerge as the result of future research, we believe that a number of poststroke depressive disorders are likely to be the result of specific changes in brain pathology and neurophysiology. Nevertheless, there are relatively few hypotheses about the pathophysiology of poststroke depression. This paper, therefore, proposes a new hypothesis for poststroke depression involving increased production of proinflammatory cytokines resulting from brain ischemia in cerebral areas linked to the pathogenesis of mood disorders. This paper reviews the evidence supporting the hypothesis that proinflammatory cytokines are involved in the occurrence of stroke as well as mood disorders linked to the brain damage. The increased production of proinflammatory cytokines such as IL-1b, TNF-a or IL-18 resulting from stroke may lead to an amplification of the inflammatory process, particularly in limbic areas, and widespread activation of indoleamine 2,3-dioxygenase (IDO) and subsequently to depletion of serotonin in paralimbic regions such as the ventral lateral frontal cortex, polar temporal cortex and basal ganglia. The resultant physiological dysfunction may lead to poststroke depression. Future investigations may explore this hypothesis through more extensive studies on the role of proinflammatory cytokines, such as IL-1b, TNF-a or even IL-18, in patients with poststroke depression. Molecular Psychiatry (2006) 11, 984-991.