This project proposes to evaluate 8-12 year old boys with Disruptive Behavior Disorders. Subjects will be evaluated on a laboratory analogue task developed to measure hostile (mean, vindictive) and instrumental (purposeful, goal-directed) aggression in response to frustration during a competitive game. Hostile and instrumental aggression are among the most well-established subtypes of aggression, based on neuroanatomy, neurochemistry, animal, and human analogue models. However, this work has not been applied to clinical samples of aggressive children. The overall goal of this approach is to evaluate these aggressive responses in relation to factors identified as influencing childhood aggression. Following the identification of provocation conditions differentially responsive to instrumental and hostile aggression, aggressive responses on the analogue task will be compared to empirically derived classifications of antisocial and disruptive behavior, to a multivariate assessment of psychosocial variables identified as predictors of delinquency, to laboratory measures of inattention and impulsivity, and to measures of social information processing biases. It is hypothesized that hostile aggression, with or without concurrent instrumental aggression, will represent a deficit in modulating behavior, as evidenced by an association with problems in impulse control and an association with predictors of delinquency and social information processing biases. Instrumental aggression is hypothesized to be goal-directed and better organized than hostile aggression, and is expected to relate to ratings of planned antisocial acts (e.g., stealing). Boys high on both types of aggressive responding are hypothesized to represent a seriously disordered aggressive group with the highest overall level of impairment.