Children reported to be hyperactive in school and with behavior difficulties at home were randomly assigned to methylphenidate, behavior therapy and placebo, or behavior therapy with methylphenidate for an 8-week period. Rating scales were obtained from teachers and parents. Independent blind observers rated childrens' classroom bheavior on a weekly basis. A behavior therapy program was implemented in the home and at school. Methylphenidate dosage was individualized. Ratings of behavior deviance were significantly reduced by all treatments. However, a significant advantage for the groups receiving methylphenidate was found over the group receiving behavior therapy and placebo. No significant differences between methylphenidate alone and methylphenidate combined with behavior therapy were obtained. Global ratings of improvement done by teachers favored the combined treatment of behavior therapy and methylphenidate over behavior therapy and placebo. No differences among treatments were found in the mothers' global ratings of improvement. The results indicate that though all three treatments were effective, methylphenidate was significantly superior to behavior therapy alone.