THIS STUDY ASSESSES THE EXTENT TO WHICH INMATE SUBCULTURE IS SUBJECT TO ORGANIZATION INFLUENCES AND MODIFICATION, WITH CORRESPONDING SHORT- AND LONG-TERM CHANGES IN INMATE ATTITUDES AND BEHAVIOR, AND THE EXTENT TO WHICH PRIOR SOCIAL EXPERIENCES PREPARING INMATES TO COPE WITH PRISON SUBCULTURE VIOLENCE DETERMINES THE NATURE OF THEIR INSTITUTIONAL ADAPTATIONS. TEN COTTAGES FROM THE TRAINING SCHOOL SYSTEM OF THE YOUTH SERVICES DEPARTMENT, WITH A WIDE RANGE OF TREATMENT GOALS AND TECHNIQUES, WERE SELECTED FOR COMPARISON. PARTICIPANT OBSERVATION AND INMATE AND STAFF INTERVIEWING WERE USED AND SUPPLEMENTED BY DATA FROM OBSERVATION FIELD REPORTS, INSTITUTIONAL RECORDS, COURT REPORTS, AND PSYCHOLOGICAL EVALUATIONS OF INMATES. IN ADDITION TO MATCHING AND COMPARING COTTAGE POPULATIONS, SEVERAL STATISTICAL TECHNIQUES WERE APPLIED TO DISCOVER RELATIONSHIPS BETWEEN PARTICULAR BACKGROUND CHARACTERISTICS AND OTHER DIFFERENCES IN THE POPULATIONS OF COTTAGES. INMATE SUBCULTURES AND INMATE ATTITUDES TOWARD STAFF, OTHER INMATES, AND SUBCULTURE ADAPTATION, THE SOCIAL STRUCTURE OF THE INMATE SUBCULTURES (REGARDING INMATE VIOLENCE AND AGGRESSION, INFORMING, AND OTHER INDICATORS OF SUBCULTURAL NORMATIVE ORIENTATION) AND THE INFLUENCE OF SEX AND RACE ON SUBCULTURAL ADAPTATION ARE EXAMINED. RESULTS SHOW THAT CORRECTIONAL INSTITUTIONS USING FORMAL COLLABORATION IN A DEMOCRATIC, EGALITARIAN STRUCTURE, USING CONSISTENCY IN RESPONDING TO TROUBLESOME INCIDENTS, OFFERING DIVERSE PROGRAM METHODS, AND DISCOURAGING STAFF-INFLICTED VIOLENCE ARE LESS VIOLENT. REFERENCES, AN INDEX, AND TABULAR DATA ARE PROVIDED.