OBJECTIVE: To test the hypothesis that amantadine hydrochloride is a safe and effective treatment for behavioral disturbances--for example, hyperactivity and irritability--in children with autism. METHOD: Thirty-nine subjects (intent to treat; 5-19 years old; IQ > 35) had autism diagnosed according to DSM-IV and ICD-10 criteria using the Autism Diagnostic Interview-Revised and the Autism Diagnostic Observation Schedule-Generic. The Aberrant Behavior Checklist-Community Version (ABC-CV) and Clinical Global Impressions (CGI) scale were used as outcome variables. After a 1-week, single-blind placebo run-in, patients received a single daily dose of amantadine (2.5 mg/kg per day) or placebo for the next week, and then bid dosing (5.0 mg/kg per day) for the subsequent 3 weeks. RESULTS: When assessed on the basis of parent-rated ABC-CV ratings of irritability and hyperactivity, the mean placebo response rate was 37% versus amantadine at 47% (not significant). However, in the amantadine-treated group there were statistically significant improvements in absolute changes in clinician-rated ABC-CVs for hyperactivity (amantadine -6.4 versus placebo -2.1; p = .046) and inappropriate speech (-1.9 versus 0.4; p = .008). CGI scale ratings were higher in the amantadine group: 53% improved versus 25% (p = .076). Amantadine was well tolerated. CONCLUSIONS: Parents did not report statistically significant behavioral change with amantadine. However, clinician-rated improvements in behavioral ratings following treatment with amantadine suggest that further studies with this or other drugs acting on the glutamatergic system are warranted. The design of these and similar drug trials in children with autistic disorder must take into account the possibility of a large placebo response.