The study found that some victim characteristics significantly predicted homicide clearances, but the significance of these characteristics varied across the five logistic regression models used in the study. Consistent with previous research, the findings show that weapon type (specifically firearms) was associated with a decreased probability of homicide clearance. There was also evidence that predictors of homicide clearances changed when the analysis excluded quick clearances. The circumstances of the homicide was one of the most influential predictors of homicide clearances. Homicides that involved "expressive" circumstances (offender's primary motive is to harm or injure the victim) were significantly more likely to be cleared than those involving "instrumental" (offender's motive is to obtain property or money), gang-related, and rape/other circumstances. Some victim characteristics significantly predicted homicide clearances across several of the analyses, including the victim's sex, age, and race. In addition, the victim's prior arrest history significantly predicted homicide clearances for instrumental and gang-related homicides. There were differences in the types of factors that significantly predicted clearances of expressive, instrumental, and gang-related homicide. This supports the argument that homicide is not a single type of crime, but rather a grouping of different types of circumstances that result in death. These differences are important in understanding homicide clearances. The data was from homicides known to the Chicago Police between January 1, 1991, and December 31, 2002 (n=7,470). The dependent variable was homicide clearances. Independent variables pertained to victim characteristics and incident characteristics. The logistic regression models used involved an analysis of all homicides; analysis of homicides excluding quick clearances; and analyses of expressive, instrumental, and gang-related homicide circumstances. 2 tables, 1 figure, 3 notes, and 18 references