THIS REPORT IS THE SECOND STAGE OF A THREE-PART EVALUATION OF RESIDENTS ADMITTED TO MASSACHUSETTS HALFWAY HOUSES, INC. ON A PRERELEASE PROGRAM DURING 1974. A DESCRIPTION OF PROGRAM ACTIVITIES IS NOT INCLUDED IN THIS REPORT. OF THE 85 PERSONS WHO PARTICIPATED IN THE PRERELEASE PROGRAM, 57 (67 PERCENT) SUCCESSFULLY COMPLETED IT AND WERE RELEASED TO THE COMMUNITY, WHILE THE REMAINING 28 RESIDENTS (33 PERCENT) WERE RETURNED TO INSTITUTIONS BECAUSE OF SERIOUS ADAPTATION PROBLEMS OR VIOLATIONS OF PROGRAM REGULATIONS. COMMITMENT, BACKGROUND, AND CRIMINAL HISTORY VARIABLES WERE CHARACTERIZED ACCORDING TO A SERIES OF DIFFERENT SPLITS FOR THE COMPLETER AND NONCOMPLETER SAMPLES. THE SPLIT THAT YIELDED THE HIGHEST CHI SQUARE WAS CHOSEN. VARIABLES THAT YIELDED A STATISTICALLY SIGNIFICANT RELATIONSHIP AT THE .05 PROBABILITY LEVEL WERE SELECTED AS INDICATORS FOR DIFFERENCES BETWEEN THE TWO SAMPLES. IN COMPARING THE NONCOMPLETER TO THE COMPLETER, IT WAS FOUND THAT SIGNIFICANTLY MORE NONCOMPLETERS THAN COMPLETERS WERE SINGLES, WERE AT LEAST 7 MONTHS BEYOND THEIR PAROLE ELIGIBILITY DATA AT THE TIME OF RELEASE, HAD BEEN PREVIOUSLY PAROLED AS AN ADULT, HAD PREVIOUSLY SERVED 18 MONTHS OR MORE IN PRISON, AND HAD THREE OR MORE PRIOR INCARCERATIONS. A SUMMARY OF THESE RELATIONSHIPS IS PRESENTED IN APPENDIX I, AND THE REMAINING VARIABLES THAT DID NOT PRODUCE SIGNIFICANT RESULTS ARE DOCUMENTED IN APPENDIX II. APPENDIX III PROVIDES FORMAL DEFINITIONS OF THE VARIABLES CONSIDERED. (RCB)