This study aimed to develop a general risk assessment system for Korean offenders. In order to achieve this goal, the first, literature was reviewed on the area of risk assessment, the second, a cognitive neuro-psychological experiment was performed to explore the utility of the Psychopathy Checklist - Revised, and the third, an empirical survey investigated the validity of a pilot risk assessment tool consisted of risk factors which were selected based on the previous studies. A literature review and an experiment study analyzed various risk factors and these factors were summarized into twelve categories. Theses categories were selected by a delphi method and produced fourteen risk variables. A survey on 200 offenders presented good internal consistency and inter-rater reliability indices. In order to search for validity evidences, criminal records were collected as criterion measures. Except the number if victims, all the risk factors functioned as expected. Therefore, other variables were explored for substituting this variable. Cognitive disability, psychiatric symptoms, and brain dysfunction seemed to differentiate individual dangerousness effectively. Finally, these variables were included into a final risk assessment system as an extra risk factor. The total score of Korean offender risk assessment system became 25 and this score effectively differentiate a hight risk group with AUC of .90 produced by ROC analysis.