We retrospectively reviewed the clinical course and response to treatment of 67 patients with tardive dystonia. The age at onset ranged from 13 to 72 years without predilection to any particular age group or sex. Patients developed tardive dystonia even after relatively short duration of exposure to dopamine antagonists (21% within 1 year). Five of 42 patients withdrawn from these drugs remitted. Overall clinical improvement occurred in 52% of patients. Tetrabenazine and reserpine were most effective (greater than 50% response rate) in controlling dystonia. Anticholinergic drugs diminished dystonia in 46% of patients. In conclusion, this review supports our original concept of tardive dystonia as a subtype of tardive dyskinesia, which is quite disabling, usually persistent, and difficult to treat. Although anticholinergics and dopamine-depleting drugs frequently improved symptoms, treatment with them only rarely led to a remission or satisfactory control of symptoms. The difficulty of treating this condition necessitates reemphasis of one important observation of this study, that this condition may develop early in the course of dopamine antagonist treatment; there is no minimum period of exposure which can be considered safe. These drugs must therefore be used only for correct medical indications, and every attempt should be made to withdraw them at the first sign of dyskinesia, particularly of the dystonic type.