Patients with Lewy body disease (LBD) frequently experience visual hallucinations (VH), well-formed images perceived without the presence of real stimuli. The structural and functional brain mechanisms underlying VH in LBD are still unclear. The present review summarises the current literature on the neural correlates of VH in LBD, namely Parkinson’s disease (PD), and dementia with Lewy bodies (DLB). Following a systematic literature search, 56 neuroimaging studies of VH in PD and DLB were critically reviewed and evaluated for quality assessment. The main structural neuroimaging results on VH in LBD revealed grey matter loss in frontal areas in patients with dementia, and parietal and occipito-temporal regions in PD without dementia. Parietal and temporal hypometabolism was also reported in hallucinating PD patients. Disrupted functional connectivity was detected especially in the default mode network and fronto-parietal regions. However, evidence on structural and functional connectivity is still limited and requires further investigation. The current literature is in line with integrative models of VH suggesting a role of attention and perception deficits in the development of VH. However, despite the close relationship between VH and cognitive impairment, its associations with brain structure and function have been explored only by a limited number of studies. 2017 https://creativecommons.org/licenses/by/4.0/ Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ( http://creativecommons.org/licenses/by/4.0/ ). Patients with Lewy body disease (LBD) frequently experience visual hallucinations (VH), well-formed images perceived without the presence of real stimuli. The structural and functional brain mechanisms underlying VH in LBD are still unclear. The present review summarises the current literature on the neural correlates of VH in LBD, namely Parkinson’s disease (PD), and dementia with Lewy bodies (DLB). [ 20 ] selected five ROI within the visual pathway to investigate differences between hallucinating and non-hallucinating PD patients in the optic chiasm area, lateral geniculate nucleus, and V1 volumes and white matter microstructure features in the optic nerve and optic radiation (the latter are described in Section 3.1.3 ). These regions were selected to examine the neural bases of VH in relation to their role in processing visual information. Volumetric reductions in VH patients were reported only in the lateral geniculate nucleus [ 20 ]. In addition to the whole brain analysis described in the previous section, Meppelink et al. [ 43 ] found a significant association between the Neuropsychiatric Inventory (NPI) [ 44 ] hallucination score, and cortical thickness in right lateralised parietal regions, namely the precuneus, and superior parietal gyrus in DLB patients ( p < 0.05 corrected). 3.1.3. Diffusion Tensor Imaging Four DTI studies were found, two on PD [ 20 , 21 ], and two on DLB [ 45 , 46 ]. Three studies investigated predefined ROIs of grey or white matter [ 20 , 45 , 46 ], while only one used a whole brain approach, namely tract-based spatial statistics TBSS (using a threshold corrected for multiple comparisons) [ 21 ]. Two fMRI studies recorded brain activity during the occurrence of visual hallucinations in single cases [ 54 , 55 ]. Both patients experienced complex VH, namely seeing animals [ 54 , 55 ] and people [ 54 ] in the MRI scanner. Howard et al. [ 55 ] scanned a DLB patient in the hallucination-free state (the patient was taking risperidone), and a second time whilst he was hallucinating (seven days after risperidone was stopped). They found decreased activation in V1 and V2 in response to photic stimulation while the patient was hallucinating compared with the hallucination-free scan [ 55 ]. On the other hand, Goetz et al. [ 77 ] investigated the relationship between changes in brain perfusion and hallucinations over one year in a combined group of patients with DLB and PDD. They found a negative association with left parietal regions, namely the posterior cingulate gyrus and the precuneus ( p < 0.05 cluster-level corrected) [ 77 ]. Another SPECT study performed factor analysis in order to investigate associations between There is a mismatch between a more prominent involvement of primary and association visual regions in brain metabolism and blood flow studies and a more prominent involvement of more frontal regions when studying GM volume or cortical thickness. None of these findings appears to be associated with a different burden of neuropathological changes. In fact, despite the association between Lewy body pathology and VH in medial temporal lobe areas [ 11 , 12 , 13 ], substantial structural alterations in these regions have not emerged from this review. We can, therefore, speculate that VH in LBD emerge only in the presence of a double hit—i.e., concomitant alterations of large functional and structural attentional networks—of which frontal lobe atrophy may be a surrogate marker, and dysfunction of visual information processing, of which occipital-temporal and parietal hypometabolism is the functional hallmark. Large attentional networks may be impaired by diffuse cortical deposition of synuclein, and even amyloid. The cause of reduction in metabolism in posterior brain regions—i.e., which crucial cortical or subcortical projections are deafferenting the occipital cortex—remains still unexplained. brainsci-07-00084-t001_Table 1 Table 1 Summary of the most consistent findings associated with VHs in LBD. Brain Regions GM Volume Functional Connectivity Task-Related BOLD Activation Glucose Metabolism Brain Perfusion Frontal ↓ ↑ ↑↓ ↓↑ Parietal ↓ ↑ ↓↑ Temporal ↓↑ Occipito- temporal ↓ ↓↑ ↓ Occipital ↓↑ ↓ BOLD: blood-oxygenation level-dependent; GM: grey matter; LBD: Lewy body disease; VH: visual hallucinations; ↓: decrease; ↑: increase.