GOALS OF THE INTENSIVE PRISON WELFARE TREATMENT EXPERIMENT FOR SHORT-TERM PRISONERS INCLUDED THE FOLLOWING: CHANGING THE TREATMENT GIVEN TO EXPERIMENTAL SUBJECTS FROM PRACTICAL HELP TO CASE WORK DISCUSSION WHENEVER APPROPRIATE; INVESTIGATING CONSEQUENCES OF THE USE OF AFTERCARE FACILITIES WHEN PRISONERS HAD RECEIVED INTENSIVE PRISON WELFARE TREATMENT; AND ASSESSING THE IMPACT OF AFTERCARE FACILITIES ON THE RESOLUTION OF PERSONAL PROBLEMS. SOCIAL AND CRIMINAL HISTORIES, PERSONALITY DATA, POSTRELEASE INFORMATION, AND DATA FROM WELFARE CASE NOTES AND SUPERVISION RECORDS WERE COLLECTED AT THE PRISON IN LIVERPOOL. THREE TYPES OF OUTCOME CRITERIA EMPLOYED WERE MEASURES OF CHANGE DURING THE PRISON WELFARE TREATMENT PERIOD, THE RATE AT WHICH AFTERCARE WAS USED, AND THE INCIDENCE OF FURTHER CONVICTION WITHIN A YEAR OF RELEASE FROM PRISON. PARTICIPANTS IN THE CONTROL GROUP WERE ABLE TO HAVE AS MUCH OR AS LITTLE CONTACT AS DESIRED WITH THE PRISON WELFARE DEPARTMENT, WHEREAS EXPERIMENTAL SUBJECTS WERE GIVEN EXTRA PRISON WELFARE CONTACTS. FOR EXPERIMENTAL AND CONTROL GROUP PARTICIPANTS WHO COMPLETED TWO PERSONALITY QUESTIONNAIRES, AVERAGE MEAN SCORES FOR PROBLEM AREAS WERE GENERALLY LARGER AT THE BEGINNING THAN AT THE END OF TREATMENT. THERE WAS NO GREATER PROBLEM REDUCTION FOR THE EXPERIMENTAL GROUP THAN FOR THE CONTROL GROUP. AFTERCARE WAS SOUGHT BY 53 PERCENT OF THE EXPERIMENTAL SUBJECTS AS COMPARED TO 47 PERCENT OF THE CONTROL PARTICIPANTS. AT THE END OF A YEAR, DIFFERENCES BETWEEN RECONVICTION RATES FOR BOTH THE EXPERIMENTAL AND CONTROL GROUPS WERE NOT STATISTICALLY SIGNIFICANT. APPENDIXES PROVIDE INFORMATION ON EXPERIMENTAL VARIABLES, THE STUDY SAMPLE, AND AFTERCARE. REFERENCES ARE INCLUDED. (DEP)