AN ANALYSIS OF THE 1973 SAMPLE (N=100) OF FURLOUGHS FROM GROUP TREATMENTS PROGRAMS SHOWED THAT THE AVERAGE COMMITMENT LENGTH FOR THE TOTAL SAMPLE WAS 6.8 MONTHS. YOUTHS WHO HAD BEEN COMMITTED PREVIOUSLY (17 PERCENT) AVERAGED SLIGHTLY SHORTER STAYS, WHILE THOSE WHO WERE TRANSFERRED DURING THEIR COMMITMENT (34 PERCENT) SERVED CONSIDERABLY LONGER, AVERAGING 9.5 MONTHS COMMITTED. TWENTY-THREE PERCENT OF THE SAMPLE RAN AWAY FROM THE PROGRAM DURING THEIR COMMITMENT, AND 7 PERCENT WERE TEMPORARILY REMOVED TO A DETENTION FACILITY DUE TO POOR ADJUSTMENT; HOWEVER, NONE OF THESE MEASURES OF INPROGRAM EXPERIENCE WAS FOUND TO BE SIGNIFICANTLY RELATED TO POSTPROGRAM SUCCESS OR FAILURE. RECIDIVISM SERVED AS THE PRIMARY MEASURE OF POSTPROGRAM OUTCOME. AT 18 MONTHS AFTER FURLOUGH, THE RECIDIVISM RATE WAS 41 PERCENT; WHEN THE FOLLOWUP PERIOD WAS EXTENDED TO A PERIOD VARYING FROM 18-30 MONTHS, THE RECIDIVISM RATE WAS 47 PERCENT. MEASURES OF 'PARTIAL SUCCESS' WERE EXAMINED IN ORDER TO SUPPLEMENT THE POSTPROGRAM OUTCOME INFORMATION PROVIDED BY THE DICHOTOMOUS RECIDIVISM DATA. FOR ONE SUCH MEASURE, THE COMMITMENT OFFENSE AND THE RECIDIVISM OFFENSE OF EACH YOUTH WHO FAILED WERE COMPARED FOR SERIOUSNESS. RECIDIVISM WAS NOT FOUND TO BE SIGNIFICANTLY RELATED TO ANY OF THE BACKGROUND VARIABLES, DELINQUENCY RECORD VARIABLES, OR INPROGRAM VARIABLES. SPECIFIC IMPROVEMENTS OF EVALUATIVE MEASURES OF PROGRAM OUTCOME ARE SUGGESTED FOR FUTURE STUDIES. DATA DERIVED FROM THE STUDY ARE INCLUDED, AND OFFENSE CODES ARE INCLUDED IN THE APPENDIX. A BIBLIOGRAPHY IS ALSO PROVIDED.