Students (N = 814) in grades 3 through 6 at a public school were screened for behavior problems using the Revised Behavior Problem Checklist. Fifty-three selected students, representing Conduct Disorder, Attention Deficit Disorder with Hyperactivity, Conduct Disorder plus Attention Deficit Disorder with Hyperactivity, Anxiety-Withdrawal Disorder, and normal controls were administered a response perseveration task and delayed responding task (DRL). It was predicted that performance on the two tasks would vary as a function of group membership. While DRL performance did not vary between groups, findings on the response perseveration task provided support for the hypothesis of the perseverative responding for reward in Conduct Disorder.