The Grasmick et al. scale is one of the most frequently used measures in criminology. Regardless of how common the scale is used, questions remain about its dimensionality and the nature of forming a composite measure from its 24 individual components. This study examines whether a composite measure is the most effective method for using the scale with a series of analyses using different approaches to combining - or not combining - these measures. Based on data from a sample of over 1,500 college students, the results indicate that a single-factor composite of the 24 items is the least effective approach in predicting crime/deviance. These results, instead, suggest using the individual components separately, multiple factors, or a composite measure made from just three of the original items lead to significant improvements over the one-factor solution.