One hundred high-risk offenders were randomly assigned to either citizen counseling with regular probation or to regular probation only. A random sample of 50 low-risk offenders assigned to regular probation served as a comparison group. Low-risk offenders were superior to the high-risk groups on measures of behavior, personality functioning, and social competence. Citizen-counseled high-risk offenders, compared to other high-risk offenders, committed fewer and less serious criminal offenses, were employed more regularly, made better educational progress, and appeared to be somewhat less rebellious, less impulsive, and more responsible as measured on the California Psychological Inventory at the end of probation. The implications and unique features of the programming, which may limit the generality of the findings to other settings, were discussed. Potentially limiting conditions included the stringent selection and matching procedures, the high caliber of the volunteers, the severe demands imposed on the time (weekly meetings over about 10 months) and talents (crisis counseling, tutoring, advocacy, friendship, and resource brokering) of the unpaid volunteers, and the availability of vital community resources, such as vocational and educational opportunities. 3 tables, 16 references. (Author abstract)