Clinical, neuropsychological and psychopharmacological characteristics were investigated for their ability to distinguish individuals with and without involuntary movements (tardive dyskinesia), among a population of 40 out-patients with bipolar affective disorder and a history of exposure to neuroleptics and lithium. Impaired performance on a test of cognitive flexibility bore the primary association with both the presence and the severity of involuntary movements. The additional relationships identified emphasized further that individual vulnerability to involuntary movements appeared to be associated not with greater duration or dosage of treatment, but with features of the bipolar illness, including number and type of affective episodes, for which that treatment was prescribed.