Patients with prefrontal damage have difficulties in shifting cognitive set, an impaired ability to move attention between one stimulus dimension and another (extra-dimensional set-shifting) and a  failure to reverse previously acquired stimulus-reward associations (reversal learning). Growing evidence in both animal and human studies suggests that these functions are segregated within the prefrontal cortex and are differentially modulated by ascending monoaminergic inputs.  Behavioural experiments have shown set-shifting and reversal learning abnormalities in patients with schizophrenia, and deficits in set shifting in patients with bipolar disorder. These data, and the presence of similar behavioural deficits in animals with prefrontal lesions, suggest potentially separable neural mechanisms in these disorders. FMRI tasks have been devised to examine the neural basis of extra-dimensional set shifting and reversal learning, but a systematic comparison in people with psychosis has not yet been made. The successful applicant will develop two fMRI tasks based on ongoing pilot work: one that clarifies the neural basis of cognitive set shifting, and another which examines reversal learning. In years 2 and 3, the successful applicant will apply the tasks to a clinical sample of people with bipolar disorder, schizophrenia and control participants (20/group). The intention is to identify activation/connectivity profiles associated with these functions, clarify how these differ in patient groups and with a view to understanding how these deficits may relate to the aetiology of psychosis and its treatment.