This chapter describes an automated methodology for the forensic authentication of bank checks. The problem of check authentication is modeled as a two-class pattern recognition problem. Color and texture features are extracted from images of genuine and counterfeit checks. A support vector machine is utilized to determine check authenticity. Classification experiments involving a dataset of 50 bank checks yielded a detection accuracy of 99.0%. The automated methodology can be used by non-specialist personnel to detect check counterfeiting in a banking environment where large numbers of checks are handled on a daily basis.