A double-blind, controlled study tested the behavioral, affective and neurological effects of antiparkinson drug [trihexyphenidyl, benztropine or bipeniden] discontinuation. Patients were evaluated at baseline and at 2 and 4 wk. Of 24 placebo patients 9 left the study early because of adverse effects; none of the 8 patients in the antiparkinsonian group did so. The placebo group had significantly more lower extremity movements, motor agitation, hallucinations and physical complaints at 2 wk and scored significantly higher in depression at 4 wk. A sizable proportion of chronic, drug-treated schizophrenic patients appear to need antiparkinsonians for clinical stability.