Little is known of factors that, on an individual basis, confer vulnerability to the emergence of involuntary movements (tardive dyskinesia) during long-term neuroleptic treatment. In this study of 88 chronic schizophrenic inpatients, 22 variables (four demographic, 14 medication history, and four features of illness) were compared for any association(s) with the presence, by differing topographies and criteria of abnormality, and severity of involuntary movements. Irrespective of the criterion used, the presence of marked cognitive dysfunction-muteness bore a consistent and highly significant primary association with both the presence and the overall severity of orofacial dyskinesia; no such association was found in relation to the presence of limb-truncal dyskinesia. Flattening of affect was the only other variable consistently associated with the presence of orofacial movements. The reliability and prominence of the association between the presence of orofacial, but not of limb-truncal, movements and cognitive dysfunction-negative symptoms suggest that these varying topographies may not constitute a unitary syndrome. This strong association, not with indexes of neuroleptic exposure but rather with features of the illness for which that treatment was prescribed, suggests some neurologic process, more subtle than may previously have been appreciated, as a vulnerability factor of some importance. In schizophrenia it appears to be intimately related to the disease process.