To determine whether intensive probationers probably would have gone to training school without the intervention of IP, the evaluation developed a statistical model of the probability of commitment to training school, using data collected on 20 youths at the four IP sites from January 1 through June 30, 1984 (before IP began). In most cases, intensive probationers' training-school risk was much greater than the risk of juveniles placed on regular probation. The recidivism of 44 intensive probationers was compared with that of 109 juveniles on regular probation over an 8-month followup period. Sixty-four percent of intensive probationers compared to 34 percent of regular probationers recidivated. The evaluation suggests that the IP program select as clients juveniles on regular probation who have a moderate risk of recidivism, rather than a very high risk, as is currently the case. Probationers having a moderate risk may be more responsive to treatment and services in the program than are offenders on the brink of removal from the community. 3 figures, 21 notes.