["\nAbstract\nThis study used cluster analysis in 126 homicide cases based on Personality Assessment Inventory (PAI; Morey, 1999). PAI was implemented by 126 homicides and then dichotomously coded for the presence or absence of cut‐off PAI scale score in order to create criteria for analysis. These cases were input for the agglomerative hierarchical cluster using Ward's method as the clustering algorithm. The results of the analysis classified five clusters: “normal,” “antisocial,” “submissive‐depressive,” “soma‐anxiety” and “isolated.” The representative characteristics were explored and compared with five clusters. As a result, these five clusters significantly differed in various criminogenic variables, such as criminal record and imprisonment experience. But, other variables were not significantly discriminated.\n", "Journal of Investigative Psychology and Offender Profiling, Volume 17, Issue 3, Page 310-331, October 2020. "]