The current study provided further validity for the use of a measure of proactive and reactive aggression that was first developed by Little and his colleagues. Confirmatory factor analyses performed with a high-risk clinical population (N = 107) revealed that the model was a good fit to the data. Structural equation models revealed significant associations in the expected direction between the Little et al. measure and another measure of proactive and reactive aggression developed by Dodge and Coie. Further analyses revealed that the proactive aggression subscales of the Little et al. and Dodge and Coie measures were significantly related to conduct problems. However, the reactive aggression subscale of the Little et al. measure was not related to social problems while both the proactive and reactive aggression subscales of the Dodge and Coie measure were. © 2009 Springer Science+Business Media, LLC.