This paper reports tests of the factor structure of the original and Revised Expagg scales using a confirmatory factor analytic approach. Current evidence regarding the dimensionality of the Expagg scale [Campbell A et al. 1992. Aggr Behav 18:95–108] proposes that the scale measures a single bipolar instrumental-expressive social representation of aggression factor. In contrast, evidence on the Revised Expagg scale [Archer J and Haigh AM. 1997. Brit J Soc Psychol 36:83–105] suggests two unidimensional factors of instrumental and expressive aggressive beliefs. Furthermore, Campbell et al. [1999. Aggr Behav 25:435–444] claim that there is equal evidence for the Revised Expagg being either one- or two-dimensional. They also suggest a short modified version of the Revised Expagg. Current evidence, however, is based on traditional and flawed exploratory factor analytic methods of scale assessment. This paper addresses the limitations of the previous psychometric assessment of these measures by implementing confirmatory factor analytic methods in the assessment of construct validity. It thereby evaluates the previous claims regarding the factor structure of these measures. The results suggest that none of the a priori specifications of the Expagg and Revised Expagg adequately explain the sample data. The results are discussed in terms of the significance of demonstrating sufficient construct validity of measurement instruments in psychological research.Aggr. Behav. 28:11–20, 2002. © 2002 Wiley-Liss, Inc.