PRISONERS IN TWO RESTRICTED PRISON ENVIRONMENTS WERE STUDIED. THEIR INDIVIDUAL RELATIONSHIPS WITH THE EXTERNAL ENVIRONMENT ARE DEFINED, ISOLATED, AND MEASURED. TWO GENERAL METHODS OF DATA COLLECTION ARE EMPLOYED. THE FIRST UTILIZES SELF-SUPPORT INSTRUMENTS TO GENERATE EMPIRICAL DATA WITH REFERENCES TO VARIOUS ASPECTS OF THE EXTERNAL ENVIRONMENT. THE SECOND USES AN INVESTIGATOR-GUIDED CONVERSATION TO GATHER ETHNOGRAPHIC DATA CONCERNING THE CONVICTS' PERCEPTIONS OF LIVING IN A LONG-TERM, MAXIMUM SECURITY INSTITUTION. A COMPARATIVE ANALYSIS OF SCORES FROM THE CONVICTS IS PERFORMED. PROFILE DATA MEASURES SPECIFIC DIAGNOSTIC ASPECTS OF INDIVIDUAL RELATIONSHIPS TO THE EXTERNAL ENVIRONMENT, AND THE MULTIPLE AFFECT ADJECTIVE CHECKLIST AND THE SOMATIC CHECKLIST ARE USED TO MEASURE ANXIETY. THE LAST PORTION OF PROFILE DATA FOCUSES ON COMPLAINTS. IN A COMPARISON OF THE EFFECTS OF THE TWO ENVIRONMENTS, DIFFERENT REACTIONS WERE NOTED ALONG SEVERAL DIMENSIONS. THE CONVICT WHO IS SENTENCED TO THE SEGREGATION ENVIRONMENT IS FACED WITH A NARROW RANGE OF POSSIBLE ADAPTIVE RESPONSES, SOME OF WHICH ARE COMPROMISING TO HIS EXPECTATIONS. AGGRESSION LEVELS ARE HIGHER IN THE SEGREGATION SAMPLE THAN THOSE OF THE POPULATION SAMPLE. FURTHER, AGGRESSION LEVEL INCREASES AS A FUNCTION OF TIME IN THE SEGREGATION ENVIRONMENT. IT APPEARS THAT ADAPTIVE RESPONSES LEARNED IN THE SEGREGATED ENVIRONMENT MAY CARRY OVER AS THE CONVICT RETURNS TO THE POPULATION ENVIRONMENT WHERE HE MAY REMAIN MALADAPTIVE. ILLUSTRATIONS, TABLES, SAMPLE QUESTIONNAIRES, AND 36 REFERENCES ARE PROVIDED.