THE AIMS OF THE STUDY WERE THREEFOLD: (1) TO DESCRIBE HOW DRUG USE AND THE RISKS ASSOCIATED WITH IT AFFECT THE STRUCTURE OF HOMICIDE IN TERMS OF WHO IS KILLED, WHO IS THE KILLER, AND THE CIRCUMSTANCES OF DEATH; (2) TO DESCRIBE THE DIFFERENCES IN HOMICIDE EVENTS IN THE TWO CITIES AND EXPLORE WHETHER THE REGIONAL CONCEPT OF VIOLENCE IS OF IMPORTANCE IN EXPLAINING ACTUAL HOMICIDES; AND (3) TO DETERMINE WHETHER DRUG USE IS A SUFFICIENTLY IMPORTANT VARIABLE TO AFFECT HOMICIDE EVENTS UNDER DIFFERENT REGIONAL AND/OR CITY CONDITIONS. A STRATIFIED SAMPLE OF HOMICIDE VICTIMS WAS USED BECAUSE THE NUMBER OF DRUG-USER HOMICIDES WAS SUBSTANTIALLY LOWER THAN THE TOTAL NUMBER OF HOMICIDES. HOMICIDE VICTIMS IN BOTH CITIES AND IN BOTH DRUG AND NONDRUG POPULATIONS WERE MORE LIKELY TO BE MALES THAN FEMALES. THE DIFFERENCES BETWEEN THE DRUG AND NONDRUG POPULATIONS OF VICTIMS IN EACH CITY WERE SIGNIFICANT WITH RESPECT TO AGE: IN PHILADELPHIA, THE DIFFERENCE IN MEAN AGE WAS 7.5 YEARS; IN DALLAS THE DIFFERENCE WAS SIGNIFICANT, THOUGH NOT SO LARGE. NEARLY 86 PERCENT OF THE VICTIMS IN PHILADELPHIA WERE BLACK, COMPARED TO ONLY 59.4 PERCENT IN DALLAS. IN PHILADELPHIA, 54.6 PERCENT WERE SINGLE, WHILE IN DALLAS BOTH THE DRUG AND NONDRUG VICTIMS WERE MORE LIKELY TO BE MARRIED. DRUG VICTIMS IN BOTH CITIES WERE MOST OFTEN FOUND DEAD ON THE STREET OR A PARKING LOT. IN BOTH CITIES, DRUG AND NONDRUG POPULATIONS SHOWED A HIGH PERCENTAGE OF CASES IN WHICH THE KILLER WAS UNKNOWN TO THE VICTIM OR THE RELATIONSHIP COULD NOT BE DETERMINED. IN PHILADELPHIA, DRUG-RELATED ARGUMENTS RANKED THIRD AS THE HOMICIDE CAUSE AMONG DRUG-USING VICTIMS. IN DALLAS, DRUG-RELATED ARGUMENTS EQUALED ROBBERY-RELATED KILLINGS. IT IS CONCLUDED THAT DRUG USE DOES AFFECT SIGNIFICANTLY THE CIRCUMSTANCES OF HOMICIDE DEATHS. TABULAR DATA, NOTES, AND REFERENCES ARE PROVIDED (BAC)