The aim of this master thesis is to find out how indexing of fiction works. I look at the consistency when users’ choices of index words are compared with each other and I also compare users’ choices to the current indexing made by libraries. I examine the subject heading list, Att indexera skönlitteratur, to find out how well it works in this particular context. The investigation included seven users who, in three steps, are creating a subject heading index of the Swedish novel Klippdockan by Cecilia Dahlheim. On the basis of their choices the results may either indicate that the consistency is good enough for subject heading index fiction, or if it is just a waste of time depending on how it is done today. The results show that consistency is low overall and that certain aspects point out changes that can be made to make subject heading index for fiction more efficient and effective. The survey shows that consistency is higher if the document text is short, if a subject heading list is used and if the words are not too specific. Today there is an increase in the amount of fiction held by libraries and there is a need for more user-friendly systems as users increasingly search in the catalogues themselves. The conclusions are that there is a need for a subject heading index for fiction in our libraries, and that there are ways to improve today's situation. / Uppsatsnivå: D