WIDESPREAD DISILLUSIONMENT WITH THE JUVENILE COURT SYSTEM, RECENT LABELING THEORIES, AND DIVERSIONARY PRACTICES EXISTS BECAUSE JUVENILE DELINQUENCY HAS INCREASED, PARTICULARLY IN SUBURBIA. PROGRAM EVALUATION HAS HERETOFORE BEEN OF SUCH POOR QUALITY THAT EFFECTIVENESS CANNOT BE MEASURED ON THE BASIS OF AVAILABLE DATA. SPECIFIC RECOMMENDATIONS ARE ENUMERATED TO IMPROVE RESEARCH TECHNIQUES MEASURING PREVENTION AND CONTROL EFFORTS. RESEARCHERS AND PROGRAM PLANNERS MUST DEVELOP BETTER STRATEGIES OF EXPERIMENTATION ' AND HYPOTHESES-TESTING AT THE OUTSET OF NEW PROGRAMS. EXPECTATIONS SHOULD BE THAT NEW PROGRAMS WILL BE UNSUCCESSFUL, AND THE RESULTS OF ALL PROJECTS SHOULD BE PUBLISHED. THEORETICAL BASES FOR NEW PROGRAMS SHOULD BE SPECIFIED, AND INDIVIDUAL AND COMMUNITY-ORIENTED THEORIES OF DELINQUENCY INTEGRATED. MOREOVER, SPECIFICITY IS REQUIRED FOR IDENTIFYING THE PRECISE NATURE AND EXTENT OF SERVICE DELIVERY, AS WELL AS THE NATURE OF THE CLIENT POPULATION. IN ASSESSING PROGRAM IMPACT, TECHNIQUES SUCH AS COST ANALYSIS AND COMPARISON, OPERATIONS RESEARCH, SYSTEMS ANALYSIS, AND NONEXPERIMENTAL DESIGNS SHOULD BE USED IN ADDITION TO TRUE EXPERIMENTAL DESIGNS. SELF-REPORT SURVEYS AND GOAL ATTAINMENT SCALING TECHNIQUES ARE SENSITIVE MEASURES OF PROGRAM SUCCESS TO BE EMPLOYED TOGETHER WITH RECIDIVISM RATES. PROGRAM COORDINATION AND PLANNING OF RESEARCH AND EVALUATION ON THE NATIONAL LEVEL ARE NEEDED. ALL NEW DELINQUENCY PREVENTION PROGRAMS SHOULD STRIVE TO PROTECT THE DUE PROCESS RIGHTS OF JUVENILE OFFENDERS, AND A PERCEIVED NEED FOR TREATMENT SHOULD NOT BE THE SOLE REASON FOR ADJUDICATION AND INCARCERATION OF YOUNG PEOPLE. REFERENCES ARE INCLUDED. (MRK)