Reduction of neuronal activity in frontocortical and limbic circuits is considered a characteristic of depression. We aimed to test this hypothesis by pooling all available data from experimental literature. All investigations were included comparing regional cerebral blood flow (rCBF) or glucose metabolism (rCMRGlc) between acutely depressed unipolar major depressive patients and healthy controls. For cortical and subcortical regions we computed the percentage difference between depressives (n = 337) and controls (n = 321). In patients with unipolar major depression rCBF and rCMRGlc were lowered in left (-4.4%, P = 0.022) and right frontal (-3.2%, P = 0.053), left (-1.7%, P = 0.061) and right temporal (-3.0%, P=0.003), left (-6.5%, P = 0.002), and right parietal (-8.8%, P=0.001), and left (-6.6%, P = 0.083) and right occipital cortex (-4.2%, P = 0.02). Moreover, there were reductions in left (-6.3%, P = 0.029) and right basal ganglia (-4.8%, P = 0.002), left (-3.4%, P = 0.114) and right thalamus (-3.1%, P = 0.036), and left limbic system (-2.2%, P = 0.127). Parameters were increased by 1.0% (P = 0.714) only in the right limbic system. There were no hemispheric asymmetries (P > 0.05). Moreover, there was no indication for an anterior-posterior gradient (P > 0.05), and thus no 'hypofrontality'. In contrast to the current view, the data indicate a diffuse cortical rather than regionalized reduction of neuronal activity in unipolar major depression.