Nineteen patients with severe tardive dyskinesia were compared with 45 patients with mild tardive dyskinesia on demographic and treatment variables. Age, sex, duration and number of psychiatric hospitalizations, and amount and duration of neuroleptic treatment did not discriminate between mild and severe tardive dyskinesia. Severe tardive dyskinesia patients were treated with significantly (p less than .01) fewer neuroleptics and had significantly (p less than .05) shorter neuroleptic-free periods. Onset was more rapid in severe tardive dyskinesia patients than in mild tardive dyskinesia patients. Severe tardive dyskinesia is more likely t result from greater patient vulnerability than from heavier neuroleptic exposure. Unipolar depressed patients who are older and on higher neuroleptic dosages may be more susceptible to severe tardive dyskinesia.