Gang and nongang homicides were sampled using a random stratified approach. Analysis confirmed previous conclusions about the implications of different law enforcement policies regarding the designation of gang homicides. Adopting a more narrow motive criterion substantially reduced the number of gang homicides reported by jurisdictions confronting violent gang activity. Comparisons of rates between jurisdictions embracing different definitional approaches were clearly not valid unless motive information was available to restructure a member-defined population of incidents. Cross-city comparisons of bivariate descriptions of gang versus nongang homicide characteristics were appropriate but within certain limitations. On the one hand, the use of firearms, participant demographics, and the number of and relation between participants were among many case characteristics that appeared to be unaffected by definitional styles. On the other hand, it was determined that comparisons of drug involvement and additional violent offenses should be approached with extreme caution. Drug motives and cocaine mentions were more common in nongang homicides, but drug sales involvement was more frequent in gang homicides. Applying motive versus more general gang involvement criteria resulted in different prevalence rates. Drug aspects emerged less frequently in homicide cases designated as gang-related under the motive criterion. 8 notes and 4 tables