In this study, the risk principle was explored within the context of an experimental outcome evaluation of the effectiveness of direct supervision roles for citizen volunteers in adult probation services. The 192 probationers who agreed to participate in the study were supervised by the Ottawa probation and parole offices during 1978-81. These probationers were serving sentences of at least 6 months, were under 26 years old at intake, had been recommended by the judge for professional therapy, and had official records showing no violent incidents. Forty probationers were assigned to professional supervision, and 152 were placed in the volunteer program according to a predetermined randomized schedule. Ontario's Level of Supervision Inventory was used to classify probationers to low and high risk groups. Three measures of probation outcome were used. Official recidivism was determined through local probation records for an average of 28 months. Separate analyses were conducted with and without technical probation violations included as indicators of recidivism. A three-level probation outcome index considered favorable early case closures as well as recidivism. Results indicate that risk assessments were predictive of correctional outcomes. Higher risk cases responded to intensive supervision. 1 table and 24 references.