The analysis focused on two categories of homicide: (1) homicides where the bonds of insularity provided by primary relationships were broken to commit instrumental offenses and (2) expressive homicides within secondary relationships. Results revealed that 28 percent of the homicides between intimates had an instrumental motive and thus could be classified as deviant. The situations included bad drug deals, debts, and disputes over property. In addition, 46 percent of the homicides that involved secondary relationships had expressive motives; this proportion was much greater than expected. This category consisted of killings of nonprimary associates over matters that did not lead to material gain. These findings and further analyses demonstrated that the motive in homicide interacts with the victim-offender relationship in important and unexpected ways. Findings raise questions about the previous understanding of the effect of relationships on motives, particularly when such concepts cannot be measured independently. Findings also underscored the conclusion that understanding of homicide events depends on access to the most detailed descriptions possible. Tables, notes, and 43 references (Author abstract modified)