THE ANALYSIS OF HOMICIDE EVENTS INCLUDES CONSIDERATION OF REGION, TIME, LOCATION, METHOD, THE IMMEDIATELY-RELATED MILIEU (INCLUDING VICTIM-PRECIPITATION), RACIAL COMPONENT, AND INSTITUTIONAL RESPONSES. THE ANALYSIS OF HOMICIDE PARTICIPANTS WAS CARRIED OUT IN ACCORDANCE WITH A FOUR-FOLD TYPOLOGY OF INMATE-INMATE; INMATE-STAFF; STAFF-INMATE; AND STAFF-STAFF HOMICIDE EVENTS. THESE WERE FURTHER DISTINGUISHED BY THE INVOLVEMENT OF SINGLE OR MULTIPLE ASSAILANTS, AND BY TIME, LOCATION, METHOD, AND IMMEDIATELY-RELATED MILIEU. THE CHARACTERISTICS OF ASSAILANTS AND VICTIMS IS PRESENTED BY PROFILES OF AGE, RACE, RELIGION, MARITAL STATUS, RESIDENCE, INTELLIGENCE, AND GRADE ACHIEVEMENT, CRIMINAL AND PRISON RECORD, MILITARY RECORD, AND OCCUPATION. DESCRIPTIONS OF SELECTED PRISON HOMICIDES HAVE BEEN USED FOR ILLUSTRATION. THE ANALYSIS OF INSTITUTIONAL DATA DEALS WITH THE PRESENCE OR ABSENCE OF CERTAIN CHARACTERISTICS OF INSTITUTIONS IN RELATION TO THE PRESENCE OR ABSENCE OF ONE OR MORE HOMICIDES IN THAT INSTITUTION IN 1973. AMONG THE CHARACTERISTICS CONSIDERED ARE: SIZE AND AGE OF INSTITUTION; DENSITY OF POPULATION; HOUSING FACILITIES; EDUCATIONAL AND RACIAL COMPOSITION OF STAFF; EXISTENCE OF SPECIAL FACILITIES - EDUCATIONAL, RECREATIONAL, THERAPEUTIC, COUNSELING, ETC.; NATURE OF PRISON DISCIPLINE AND PRISON PRIVILEGES. (AUTHOR ABSTRACT)