A TOTAL OF 627 SELECTED MALE FELON INMATES RELEASED FROM CALIFORNIA PRISONS BETWEEN APRIL 1, AND DECEMBER 31, 1976, WERE RANDOMLY ASSIGNED TO EXPERIMENTAL AND CONTROL GROUPS (310 IN THE EXPERIMENTAL GROUP AND 317 IN THE CONTROL GROUP) EITHER DIRECTLY AT RELEASE OR ON A DELAYED BASIS 3-4 MONTHS AFTER RELEASE. EXCLUDED FROM PARTICIPATION WERE 62 PERCENT OF THOSE INMATES NORMALLY PAROLED (INMATES COMMITTED FOR FIRST DEGREE MURDER OR A SEX OFFENSE, AND INMATES PAROLED UNDER SPECIAL CONDITIONS). THE EXPERIMENTAL GROUP WAS GIVEN SUMMARY PAROLE SUPERVISION, WHICH PROVIDED THAT PAROLE AGENTS WERE TO INITIATE CONTACTS ONLY IF RETURN TO CRIMINAL ACTIVITY WAS KNOWN OR SUSPECTED; SERVICES WERE TO BE PROVIDED ONLY AT THE REQUEST OF THE PAROLEE. IN THE COURSE OF THE STUDY, CASES UNDER SUMMARY SUPERVISION EXPERIENCED A MEDIAN OF 5 CONTACTS OVER A 6-MONTH PERIOD WHILE CASES UNDER REGULAR SUPERVISION HAD 10 CONTACTS OVER THE SAME PERIOD. EXPERIMENTALS AND CONTROLS WERE DETERMINED TO HAVE NO SIGNIFICANT BACKGROUND DIFFERENCES WHICH WOULD INVALIDATE THE DISTINCTIVE IMPACT OF THE NATURE OF PAROLE SUPERVISION ON PAROLE OUTCOMES. THE TWO GROUPS WERE COMPARED BY MEASURES OF FREQUENCY AND SEVERITY OF KNOWN CRIMINAL ACTIVITY. OVERALL THERE WERE NO SIGNIFICANT DIFFERENCES IN PAROLE OUTCOME BETWEEN THE TWO GROUPS AFTER 6 AND 12 MONTHS IN THE COMMUNITY. A MORE SERIOUS DEGREE OF CRIMINAL ACTIVITY WAS FOUND IN THE SUMMARY PAROLE GROUP AFTER 1 YEAR, ALTHOUGH THESE FINDINGS WERE BASED ON SMALL CELL SIZES. BASED ON THESE FINDINGS, A REDUCED LEVEL OF SUPERVISION CAN BE IMPLEMENTED FOR THE 38 PERCENT OF THE PAROLE POPULATION INCLUDED IN THIS PROJECT. TABULAR AND GRAPHIC DATA ARE PROVIDED, AND THE APPENDIXES PROVIDE INFORMATION ON THE INSTRUMENTS AND PROCEDURES USED IN THE STUDY. (RCB)