This PhD project starts from different theoretical models dealing with youth justice (welfare, justice, sanction, restorative, risk-management model). The central research question is: To what extent elements of these models can be found in the day-to-day practice of Belgian youth court judges?. In other words, can the daily practice of Belgian youth court justice be described in terms of one or more model(s)? What is their practical relevance? The central research question consists of four specific research questions: (1) Which are the different theoretical youth justice models? What are their key characteristics?; (2) On which model(s) the current Belgian legislation is based on?; (3) By which model(s) the daily practice of youth court judges is characterized?; (4) Is one model more dominant or are there differences in relation to the case, the judge and/or the context? These questions are answered by a literature review, an in-depth study of the legal framework, as well as an empirical research. In four judicial districts two either in the Flemish and French speaking part of Belgium judges are observed and interviewed about their decision-making process in real cases (80 cases in total). For each case, we examine the file, we observe the court room hearing, we interview the judge about the case and his/her decision and finally, we analyse the (written) judgement. In addition, each judge is invited to a more general interview relating to, among others, characteristics of the judge (background, previous function, role perception, etc.) and context factors (culture at the youth court, formal / informal cooperation with others, perceived role of other parties involved, time pressure, case load, etc.). We are primarily using qualitative research methods that allow us to gain an insight into the views of youth court judges. The central focus is on the judge, as an actor, without however losing sight of the wider (structural, cultural) context in which the judge operates and which also has an impact.