||The impact study of the Youth Offender Demonstration Project (YODP) was intended to be an ambitious random assignment evaluation that would address questions about how to work most effectively with youth offenders to prevent recidivism and increase their employment and earnings. To answer these questions, the study proposed to have courts randomly assign youth offenders at conviction to incarceration, an alternative sentencing program, or incarceration with intensive aftercare. However, the research team was unable to persuade the courts and YODP programs in any of the six selected jurisdictions to accept the intended random assignment design. The report describes how the Employment and Training Administration and its contractor, Social Policy Research Associates, attempted to persuade the courts and YODP grantees to accept a random assignment study design and discusses the reasons why these efforts were unsuccessful. The report provides valuable recommendations, from this experience and other impact studies, for government agencies and researchers when attempting random assignment studies.