One thing we know for certain after decades of functional imaging in schizophrenia is that it is not a disorder that can simply be attributed to circumscribed lesions in the brain. It is, in other words, a disorder of the connectivity of the brain. In this overview, we will consider the power of connectivity analyses of functional MRI (and PET) data as tools for translational neuroscience. We describe the patterns of functional and effective disconnectivity seen in schizophrenia and particular psychotic symptoms, those that appear to be attributable to genetic and/or environmental risk factors for psychosis, the potential of these disconnectivities as trait and state biomarkers, and their sensitivity to drug effects. We conclude that substantial work needs to be done on standardising connectivity analyses across laboratories and that disconnectivity studies should be an integral part of drug discovery programmes.