THE PRISONS COMPARED ARE THE DE BOSCHPOORT PRISON IN BREDA, WHICH MAINTAINS AND INDIVIDUAL CELL LIVING ARRANGEMENT AND TIGHT SECURITY, AND THE NEDERHEIDE PRISON IN DOETINCHEM, WHICH PROVIDES GROUP LIVING ARRANGEMENTS, EDUCATIONAL PROGRAMS, AND DEGREE OF FREEDOM OF MOVEMENT. THE DE BOSCHPOORT SAMPLE, WHICH IS SMALL, IS SUPPLEMENTED BY SUBJECTS FROM A SIMILAR REGIME AT ARNHEM PRISON. SAMPLES STUDIED CONSISTED OF THE PRISONERS RELEASED DURING THE PERIOD 1971 TO 1973 FROM NEDERHEIDE (535), IN 1971 AND 1972 FROM ARNHEM (351), AND IN 1973 FROM DE BOSCHPOORT (212). SUBJECTS ARE CONSIDERED RECIDIVISTS IF THEY ARE CHARGED WITH OFFENSES WITHIN 2 YEARS OF RELEASE. IN CALCULATING THE EFFECTS OF THE PRISON REGIMES, OTHER FACTORS WITH POSSIBLE INFLUENCE ON RECIDIVISM, SUCH AS AGE AND CRIMINAL RECORD, ARE CONSIDERED. RESULTS SUGGEST THAT PRISON REGIMES HAVE NO APPRECIABLE INFLUENCE ON RECIDIVISM OF CONVICTS LITTLE DISPOSED TO RECIDIVISM, BUT THAT THE REGIME AT NEDERHEIDE HAS RELATIVELY POSITIVE EFFECTS ON RECIDIVISM OF PRISONERS STRONGLY INCLINED TO RECIDIVATE. NO CONNECTION APPEARS TO EXIST BETWEEN THE PRISON REGIMES AND THE SPEED OR SERIOUSNESS OF RELEASED CONVICTS' RECIDIVISM. AS FOR EFFECTS OF THE PRISON REGIMES ON SPECIFIC PRISONER GROUPS, THE NEDERHEIDE REGIME IS MORE LIKELY THAN THE DE BOSCHPOORT REGIME TO REDUCE RECIDIVISM RATES OF PRISONERS UNDER AGE 35 WITH A RECENT CRIMINAL RECORD, RELATIVELY FREQUENT CONVICTIONS BEFORE IMPRISONMENT, AND AN UNCONDITIONAL PRISON TERM OF MANY MONTHS. HOWEVER, THERE ARE NO DIFFERENCES IN EFFECTS OF THE TWO PRISON REGIMES ON PRISONERS OVER 35 YEARS OLD WITH LESS RECENT CRIMINAL RECORDS, FEW CONVICTIONS BEFORE THE MOST RECENT IMPRISONMENT, AND AN UNCONDITIONAL PRISON SENTENCE OF ONLY A FEW MONTHS. THE RESULTS OF THIS STUDY SHOULD BE CONSIDERED IN PLACING CONVICTS IN PRISONS, ALTHOUGH PLACEMENT OF MANY PRISONERS INCLINED TO RECIDIVISM IN DOETINCHEM MIGHT CHANGE THE NATURE OF THE PRISON. TABLES, APPENDIXES, AND NOTES ARE SUPPLIED.