The hodotopic framework is a recent revision of Geschwind's disconnection paradigm incorporating advances in functional and white matter imaging. Its intention is to help clinico-pathological correlations across a range of neurological and psychiatric conditions and generate novel research questions. Here I consider hallucinations within this framework. The paper is divided into three parts. The first reviews the auditory and visual hallucination literature from the dual perspectives of dysfunction localised to specific brain regions (topological) and dysfunction related to connections between brain regions (hodological), combining evidence from tractography, functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) studies. Patients prone to hallucinations have complex, task-specific hodological abnormalities that persist between hallucination episodes. During hallucinations, topological increases in activity are found whose location defines hallucination content and modality. Whether these activity increases are accompanied by transient hodological change is unclear. The second part of the paper addresses this issue in EEG and fMRI studies of a 200-year-old paradigm. Photic stimulation within a specific frequency and luminance range induces hallucinations of geometrical patterns, colours and motion in normal subjects. By comparing hallucination-inducing with control stimulation, topological activity increases were identified in visual areas whose specialisations matched the induced hallucination contents. During hallucinations, fMRI connectivity between LGN and cortex changed from a positive to negative relationship while EEG connectivity between occipital and other brain regions increased. The complex and dynamic topological and hodological changes during induced hallucinations are consistent with a shift in thalamocortical circuitry from tonic to burst mode and may have direct relevance to the Charles Bonnet Syndrome. The third part of the paper considers the relevance of the finding to other disorders, examines the strengths and limitations of our current imaging approaches to connectivity and looks to future developments in the field.