THE MASSACHUSETTS COMMUNITY ASSISTANCE PROGRAM (MASSCAP) ASSIGNED VOLUNTEER COMMUNITY ASSISTANTS TO SPECIALLY LABELED OFFENDERS PAROLED FROM THE MASSACHUSETTS CORRECTIONS SYSTEM. THE PROGRAM WAS MODELED ON CITIZEN ADVOCATE PROGRAMS DEVELOPED FOR MENTALLY RETARDED PERSONS. THE EVALUATION CONSISTED OF A QUASI-EXPERIMENTAL DESIGN IN WHICH 37 PARTICIPANTS IN THE DEMONSTRATION PROJECT WERE COMPARED WITH A CONTROL GROUP OF 37 SIMILAR SPECIALLY LABELED OFFENDERS RELEASED UNDER TRADITIONAL PAROLE ARRANGEMENTS. THE STUDY HYPOTHESIS WAS THAT COMMUNITY ASSISTANTS WOULD SIGNIFICANTLY IMPROVE THE PAROLE SUCCESS OF SPECIALLY LABELED OFFENDERS. EXPERIMENTAL AND CONTROL GROUP MEMBERS RECEIVED SIMILAR SCORES ON THE ADAPTIVE BEHAVIOR SCALES OF THE AMERICAN ASSOCIATION ON MENTAL DEFICIENCY AND SHOWED SIMILAR ATTRIBUTES, SOCIAL CHARACTERISTICS, AND PAST EXPERIENCES, INDICATING THAT THE TWO GROUPS WERE EQUIVALENT, BASED ON THESE MEASURES. EVALUATION OF THE PAROLE EXPERIENCE OF THE EXPERIMENTAL AND CONTROL GROUPS INDICATED NO DIFFERENCES, SUGGESTING THAT, IN TERMS OF THE MEASURE USED HERE (NUMBER OF DAYS ON PAROLE AND PAROLE SUCCESS), COMMUNITY ASSISTANTS HAD NO SIGNIFICANT IMPACT ON SPECIALLY LABELED OFFENDERS. IN A REVIEW OF THE ORIGINAL ADVOCACY MODEL AND THE MASSCAP PROGRAM, THE AUTHOR CONTENDS THAT THE PROGRAM COULD PROVE MORE EFFECTIVE IF IT IS MODIFIED AND STRENGTHENED TO GIVE CONSIDERATION TO THE SPECIAL CHARACTERISTICS AND PROBLEMS OF SPECIALLY LABELED OFFENDERS. (AUTHOR ABSTRACT MODIFIED)