The results indicate that homicides follow a meaningful structure that can be interpreted along a two-dimension model: a business-personal dimension and a settlement-escalation/angry brawl dimension. The business-personal dimension involves homicides ranging from business and organized crime killings to homicides caused by family issues. The settlement-escalation/angry brawl dimension involves homicides that range from planned and distant executions to angry killings that take place during a fight. The findings suggest that instead of grouping homicides according to distinct categories, they should be ranked along a circular structure and take into account the continuum of homicide incidents. The advantage of employing dimensions instead of categories is that the “fuzzy” homicides that fail to fall into distinct categories can be easily accommodated along the dimension continuum. Data were gathered on all 202 homicide incidents that occurred in the Netherlands during 1998 from police databases, newspaper clippings, and a database containing information on all prosecution and court decisions. Data were analyzed using multivariate statistical models. Future research should ensure that the empirical classification system employed by the current study not diffuse the classification of homicides. Tables, figures, appendixes, notes, references