We present an overview of the types of imaging experiments that can be performed on psychologically impaired patients. The critical observation from such studies is a differential pattern of activation in the patients and normals. Underactivity is interpretable only when the patients make normal responses. In this context, a failure to activate a component region of the normal system implies that this region was not necessary for task performance. Overactivity indicates either cognitive or neuronal reorganisation. Neuronal reorganisation is indicated only if the patient performs the task using the same set of cognitive operations as normal subjects. Cognitive reorganisation can be demonstrated if the same activation pattern is elicited by normals when they are co-erced into using the same cognitive implementation as the patient. We conclude that the interpretation of neuroimaging studies of psychologically impaired patients depends on intact task performance and a detailed task analysis. When these criteria are met, patient studies can be used to identify: (1) necessary and sufficient brain systems, (2) dysfunction at sites distant to damage, (3) peri-damage activation, and (4) compensation either at a neuronal level when pre-existing cognitive strategies are re-instantiated using duplicated neuronal systems (degeneracy), or at a cognitive level when alternative cognitive strategies (and their corresponding brain systems) are adopted.