THE PRIMARY GOAL OF THE DSO PROGRAM WAS TO REMOVE PRESENTLY INCARCERATED STATUS OFFENDERS FROM INSTITUTIONS AND TO PREVENT FUTURE INCARCERATION OF STATUS OFFENDERS ENTIRELY. THIS EVALUATION WAS HAMPERED BY SUCH PROBLEMS AS DIFFICULTY IN GATHERING DATA REGARDING SITE SPECIFICATIONS OTHER THAN CONNECTICUT'S, DELAYS IN PROGRAM IMPLEMENTATION, AND DEPARTURE FROM THE ORIGINAL RANDOMIZED-ENTRY PROGRAM DESIGN. IT WAS CONCLUDED THAT INSTEAD OF A DSO PROGRAM TO CREATE NEW ALTERNATIVES FOR STATUS OFFENDERS, IT MIGHT HAVE BEEN JUST AS FEASIBLE AND LESS EXPENSIVE TO HAVE RESTRICTED THE USE OF DETENTION AND TO HAVE SECURED COMMITMENT FOR STATUS OFFENSES WITHOUT PROVIDING ANY NEW ALTERNATIVES. DIFFERENT DATA SOURCES PRODUCED CONFLICTING FINDINGS REGARDING THE PATTERN OF DETENTIONS AND INSTITUTIONAL COURT COMMITMENTS BEFORE AND AFTER THE START OF THE DSO PROGRAM, SO THAT IMPACT IN THIS REGARD WAS DIFFICULT TO ASSESS. OF THE PURE STATUS OFFENDERS DETAINED DURING THE PROGRAM PERIOD, ABOUT 60 PERCENT WERE ELIGIBLE FOR THE DSO PROGRAM; OF THESE, ABOUT 60 PERCENT ENTERED. THE TYPICAL DSO CLIENT WAS A WHITE, 14- OR 15-YEAR-OLD GIRL, BROUGHT TO DETENTION BY POLICE ON RUNAWAY CHARGES; THIS WAS USUALLY HER FIRST COURT REFERRAL, SHE WAS PROBABLY ATTENDING SCHOOL FULL TIME AND LIVING WITH HER FAMILY WHICH WAS OFTEN NOT INTACT. THE PREPROGRAM GROUP AND THE THREE DSO GROUPS CAN BE SEEN AS FOUR TREATMENT MODELS: COMMUNITY-BASED, MINIMUM INTERVENTION; COURT-BASED, MINIMUM INTERVENTION; COMMUNITY-BASED, MAXIMUM INTERVENTION; AND COURT-BASED, MAXIMUM INTERVENTION. COMPARISON ACROSS ALL GROUPS SHOWED THAT THERE WERE NO SIGNIFICANT DIFFERENCES IN OUTCOME BETWEEN THE FOUR TREATMENT MODELS. IT IS SUGGESTED THAT FOR THE DEINSTITUTIONALIZATION EFFORT TO ACHIEVE ITS GOAL IT MUST HAVE THE COURAGE OF ITS CONVICTIONS AND RECOGNIZE THE NECESSARY DEPENDENCE OF DEINSTITUTIONALIZATION ON DECRIMINALIZATION. FOOTNOTES, TABLES, FIGURES, AND AN APPENDIX ARE PROVIDED. (LWM)