"The genesis of a policy involves the recognition of a problem. What counts as a problem and how a problem is defined depends upon the way in which policy makers seek to address an issue or an event." (Parsons 1995 p 87)The policy problems that are brought into educational policy making are often very complex. How a problem is defined and conceptualised largely influences what is seen as the solutions available and therefore affects the shape and content of a particular education policy. Furthermore, successful implementation often depends on achieving an "accurate" description and definition of the problem that the policy addresses.In 1999, The Swedish Parliament decided to introduce an experiment that permits compulsory schools to leave the regulations of the national time schedule. The participating schools enjoy complete autonomy in e.g. setting the number of teaching hours in every subject. Swedish compulsory schools clearly are the most autonomous in regards to time allocation compared o other European countries, and this is the case even when the trial not is taken into account.More generally, Swedish educational policy making since the 1970s has taken the direction of decentralisation, deregulation and market orientation. These efforts have been clearly demonstrated in particular in the education policies from the 1990s. Contemporary policy making on the one hand includes efforts to further decentralisation. The ongoing experiment with local time schedules is one example of this. On the other hand, other policies have been adopted that strengthen the responsibilities of the state. Against such a background, why introduce the experiment and what is the experiment going to accomplish? The aim of the paper is to explore and analyse the representation(s) of the policy problem that the experiment with local time schedules is set to solve.The theoretical point of departure is a post-positivist approach in policy analysis, originating from Carol Lee Bacchi (1999), asking What's the Problem (represented to be)? The basic line in Bacchi's approach is that every policy or policy proposal in itself contains a diagnose of the problem. This is referred to as the representation of the problem. The decision to introduce an experiment with local time schedules includes implicit or explicit understandings (and hence representations) of the problem that the policy proposal is going to address. Drawing on official documents and interviews, the task is to uncover and contextualise these representations, and asking what remains untouched. More specifically, the following two questions will be highlighted: What is the problem represented to be, in the case of the Swedish experiment with local time schedules? What presuppositions or assumptions underlie this representation and what is left unproblematic in this representation? Four representations of the policy problem that the experiment is set to solve are identified. One representation includes an expressed need for something to happen, to break off stagnation and a wish for a (desired) change to take place. The problem is also represented as a dysfunctional way of steering Swedish education, where the national time schedule in fact already is an obsolete governing mechanism. The problem is additionally portrayed as lack of individualisation, and as an opportunity to put the pupils' needs on top of the agenda. Last, the problem that the experiment is going to address is also represented as originating from an effort to strengthen management by objectives and decentralisation.