Previous studies have demonstrated that severity of language impairment is a critical determinant of long-term outcome for schizophrenic children in the age group of two to six years. Language impairment was therefore used as the major measure of severity in this study, designed to test whether subgroups of schizophrenic children, divided according to prognostic criteria, would respond differently to drugs as well as to placebo. Responses to both nonpharmacologic aspects of treatment and to the drug itself were found to be related to the initial severity of language impairment. Thus the major developmental defect determining long-term outcome for schizophrenic children in this age group also predicted the immediate responsiveness to treatment.Abstract Teaser