The purpose of the study was to investigate whether homicide committed and investigated within different national contexts could be usefully compared as a first step toward standardizing systems for classifying homicide. The sample consisted of 210 solved and unsolved single-offender, single-victim homicide cases from 1983 to 2000 originally collected from the Athens Police in Greece. Thirty-one variables were coded across the 210 cases to reflect the actions by the offenders on the victim, traces of behaviors left at the crime scene, and variables reflective of the characteristics of the victim. Of the total sample of 210 cases, 72 percent of the cases involved a male victim. Of the 144 cases where the gender of the offender was known, 90 percent of the cases involved a male offender. The victim ages ranged between newborn children and 87 years, with the mean age of 42 years and median age of 37 years. The most frequently occurring actions at the crime scene involved the offense occurring at night, the offender displaying extreme violence over and beyond that needed for the killing of the victim, and the offense occurring inside. Results show that the information that is contained in Greek police files can be detailed and extensive enough to allow for a comprehensive modeling of actions during homicide, and the results also confirm that behaviors are a useful unit of analysis to help distinguish between offenses and offenders in terms of their expressive and instrumental components. They further indicate that the overall behavioral structure of homicide goes beyond cultural differences between homicides in different countries, although further detailed analysis is needed before any reliable conclusions can be made. 2 figures, 3 tables, 8 notes, 28 references