Findings show that each homicide type had distinctive situational covariates that were related to the nature of the incidents. For example, the situational covariates of domestic homicides differed from those of other homicide types in incident location, the use of guns, the number of suspects, and the characteristics of victims and suspects. Drug-related homicides differed from other homicide types in their occurrence in public housing complexes during weekdays, the use of guns, and the lack of alcohol or drug impairment by victims and suspects. Dispute homicides, on the other hand, were distinct from the other types regarding the variables of drug and alcohol impairment, victim-offender relationship, their prevalence during weekends, and the race of victims. Robbery homicides differed from the other types primarily in victim-offender relationship, number of suspects involved in the offenses, and the characteristics of the victims. The most distinctive homicide type was domestic homicide, followed by robbery. Drug and dispute homicides had more similarities than differences with one another. These findings indicate the importance of studying homicides by type rather than in the aggregate. The findings can assist homicide investigators regarding motives for homicides marked by particular covariates, and they can also help in devising prevention policies for various homicide types. Suggestions are offered for future research. The primary data source was the investigation files of the homicide unit of the Newark Police Department. The analysis addresses the 659 homicides that occurred from January 1, 1997, to December 31, 2005. The dependent variable was homicide motive (why the offender killed the victim). Independent variables were location, weapon, alcohol/drug use, victim-offender relationship, and number of suspects. 7 tables, 12 notes, and 42 references