FOCUSING ON THE EFFECTIVENESS OF A 'THERAPEUTIC MILIEU' TYPE PROGRAMME IN PRODUCING LASTING BEHAVIOR AND ATTITUDE CHANGE, THE RESEARCH WAS INTEGRATED WITH THE ACTIVITIES AT THE CENTRE, INVOLVING STAFF AND RESIDENTS IN A VARIETY OF DATA COLLECTION PROCESSES. A TOTAL OF 179 OF THE SAME WOMEN WAS INTERVIEWED ONE YEAR AFTER THEIR RETURN TO THE COMMUNITY, THIS SUB-SAMPLE SHOWN TO BE HIGHLY REPRESENTATIVE OF ALL THE WOMEN INCLUDED AT INTAKE. AMONG THE TYPES OF PRE-INSTITUTIONAL DATA EXAMINED WERE: AGE AT COMMITMENT, EDUCATION, SOCIO-ECONOMIC STATUS, FAMILY HISTORY, AND PREVIOUS CRIMINAL INVOLVEMENT. POSTINSTITUTIONAL FACTORS EXAMINED INCLUDED RECIDIVISM, EMPLOYMENT, FINANCIAL ADEQUACY, PHYSICAL AND EMOTIONAL HEALTH, FAMILY RELATIONSHIPS, AND RESIDENTIAL ADEQUACY. MANY VARIABLES KNOWN AT ADMISSION WERE FOUND TO BE SIGNIFICANTLY RELATED TO SUBSEQUENT RECIDIVISM, INCLUDING PRIOR CRIMINALITY, EARLY FAMILY PROBLEMS, SERIOUS PERSONAL PROBLEMS, AND AN UNSTABLE HISTORY OF EMPLOYMENT. INSTITUTIONAL FACTORS WHICH SEEMED TO BE RELATED TO RECIDIVISM WERE LENGTH OF TIME IN THE PROGRAM, MISBEHAVIOR DURING THE PROGRAM, AND ATTITUDES OF INMATES ON RELEASE. THE MOST SIGNIFICANT POST-INSTITUTIONAL FACTOR WAS FOUND TO BE EMPLOYMENT. SEVERAL TABLES OF DATA ON THE VARIABLES AND THEIR RELATION TO RECIDIVISM ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED)