THE SAMPLE FOR THIS PROJECT CONSISTED OF 2 MATCHED GROUPS OF 29 JUVENILES EACH SELECTED FROM YOUTHS SENTENCED TO PERIODIC DETENTION IN THE WELLINGTON MAGISTRATE'S AND CHILDREN AND YOUNG PERSONS COURTS BETWEEN 1974 AND 1977. BOTH GROUPS PARTICIPATED IN THE SAME PROGRAM ADMINISTERED BY THE SAME STAFF, BUT GROUP 1 YOUTHS WERE REQUIRED TO RESIDE IN THE CENTER FRIDAY AND SATURDAY NIGHTS WHILE GROUP 2 YOUTHS WERE NOT. VARIABLES USED IN MATCHING INCLUDED SENTENCE LENGTH, DELINQUENCY HISTORY, AGE, RACE, EDUCATION, AND FAMILY BACKGROUND. YOUTHS WERE FOLLOWEDUP 1 YEAR AFTER THEIR SENTENCE ENDED TO DETERMINE IF ANY HAD BEEN RECONVICTED. SERIOUS TRAFFIC CONVICTIONS WERE INCLUDED BECAUSE PERIODIC DETENTION IS BEING USED INCREASINGLY FOR THESE OFFENSES. A PRIOR EVALUATION OF OFFENDERS SENTENCED TO PERIODIC DETENTION INDICATED A TREND TOWARD HIGHER RECONVICTION RATES AND CUSTODIAL SENTENCES UPON RECONVICTION. ANALYSIS OF THE DATA COLLECTED FOR THIS STUDY SHOWED THAT RESIDENTIAL DETAINEES DID NOT BREACH THE PERIODIC DETENTION ORDER MORE OFTEN THAN NONRESIDENTIAL DETAINEES, EVEN THOUGH THEY WERE SUBJECT TO MORE RESTRAINTS. NO DIFFERENCES BETWEEN THE GROUPS WERE DISCOVERED REGARDING SENTENCE COMPLETION RATES, RECONVICTIONS DURING THE PERIODIC DETENTION SENTENCE, AND RECONVICTIONS DURING THE FOLLOWUP PERIOD. MOST OFFENDERS FROM BOTH GROUPS WERE RECONVICTED IN THE YEAR FOLLOWING PERIODIC DETENTION AND RECEIVED CUSTODIAL SENTENCES. IN SOME CASES, THE RESIDENTIAL OFFENDER WAS GIVEN A FURTHER PERIODIC DETENTION SENTENCE, WHILE THE MATCHED PARTNER FROM THE NONRESIDENTIAL SAMPLE WAS NOT RECONVICTED OR RECEIVED A LESSER SENTENCE. THIS DIFFERENCE MIGHT BE SIGNIFICANT IN A LARGER SAMPLE. OVERALL, NEITHER THE RESIDENTIAL NOR THE NONRESIDENTIAL SETTING COULD BE REGARDED AS MORE EFFECTIVE IN PREVENTING RECONVICTION. TABLES AND THREE REFERENCES ARE INCLUDED. THE APPENDIXES CONTAIN A LIST OF TRAFFIC OFFENSES DEFINED AS SERIOUS IN THE STUDY, AN OFFENSE SERIOUSNESS SCALE, CODING DEFINITIONS, AND A DISCUSSION OF THE MATCHING TECHNIQUES.