The RAG is a recently developed assessment instrument designed to predict violent behavior of mentally disordered male offenders. This test of RAG with a nonmentally disordered offender population used 143 federally sentenced Canadian male offenders who had not been diagnosed with a mental disorder. Each subject was individually interviewed and assessed by one of two doctoral-level psychologists or a doctoral student in psychology trained in the administration of all of the psychometric measures. All of these measures were administered and verified through a review of available case file information. The findings support the hypothesis claimed by Harris et al. that the RAG can be expanded from use with exclusively mentally disordered offenders to use with non-mentally disordered offenders as well. The study also found that violent offenders were significantly more likely to score higher on the RAG than were nonviolent offenders. Seven of the 12 items distinguished these two groups of offenders. 5 tables and 50 references