This study evaluates the validity of two aggression scales for predicting observations of malicious or disruptive behavior at school. Subgroups of a sample of 1,560 children (age 8.6+/-1.5 years) were assessed using (a) peer nominations of aggression, (b) teacher reports on the Teacher Report Form (TRF) of the Child Behavior Checklist (CBCL) Aggression scale and the peer nomination items, or (c) self-reports on the peer nomination items. Criteria were observations of physical, verbal, initiated, retaliatory, malicious, and disruptive behaviors. Teacher report peer nominations predicted observed physical, verbal, initiated, and retaliatory aggression and disruptive behavior. Peer nominations predicted physical aggression, verbal aggression, initiation and disruptive behavior, and TRFs predicted verbal, initiated, and disruptive behavior. Self-reports did not significantly predict any behavior. Implications for assessment of aggression are discussed.