More than 500 juveniles were randomly assigned to the in-home programs or to a control group of State wards. The in-home programs retained about one-half of their cases in the community until successful termination at one-third the cost of commitment. The experimental and control groups showed little difference on recidivism outcomes measures two years after assignment. More charges had been filed against the in-home program participants, however this difference disappeared after controlling for amount of time "at large," while the average seriousness of the charges was higher in the control group. On measures of self-reported delinquency, the in-home program cases reported a greater decrease in violent crimes. The authors conclude that in-home programs are a cost-effective and viable alternative for many youthful offenders. 2 figures, 13 tables, 39 references. (Author abstract modified)