Data for the study derive from interviews with 30 social workers, analyses of the contents of 24 background investigations for juvenile courts, and observation of 30 discussions between juvenile court assistants and their clients as well as of 100 juvenile court proceedings. The goal of the study is to determine if labeling processes of juvenile offenders already begin when court assistants select material for written background reports to the court prior to the main trial. Results indicate that data in written reports come primarily from a single interview of the social worker with the client and that data are structured by the form used for the report. Positive and negative comments by social workers on the reports are usually middle-class oriented (e.g., praise for punctuality). Delinquent behavior is explained in reports with theories of individual psychology and common-knowledge statements. The social category of the offender, which is in the view of certain researchers the most important factor in sentencing, is established by three variables: leisure-time behavior, family situation, and academic and professional history of the accused. In court proceedings, these factors are repeatedly cited by judges and prosecutors. If information about these areas is missing from background reports, judges seek to obtain the information by directly questioning social workers or the accused. The positive or negative determination in these areas appears to be highly significant in rulings on the presence or absence of destructive inclinations as a criterion for imprisonment. Thus, both in writing reports and in giving court testimony, court assistants tend to anticipate judges' interest in these factors. At least a part of the social workers consider their work successful if judges reach the preplanned conclusions on the basis of the 'facts' emphasized in background reports. Notes and a 17-item bibliography are supplied.