Some scientists write literary fiction books in their spare time. If these books contain scientific knowledge, literary fiction becomes a mechanism of knowledge transfer. When this is the case, in the framework of the distinction of formal versus informal knowledge transfer, we conceptualise literary fiction as spare-time formal knowledge transfer. We model scientific authorship of literary fiction as a function of geographical co-location between editors and authors, and knowledge transfer as a function of the type of scientist (academic or non-academic). We distinguish between direct knowledge transfer (the book includes the scientist’s research topics), indirect knowledge transfer (scientific authors talk about their research with cultural agents) and reverse knowledge transfer (cultural agents give scientists ideas for future research). Through mixed-methods research and a sample from the Valencian Community (Spain), we find that scientific authorship accounts for a considerable percentage of all literary fiction authorship and is negatively related to geographic co-location with editors. Academic scientists do not transfer knowledge directly so often as non-academic scientists, but the former engage into indirect and reverse transfer knowledge more often than the latter. We draw propositions about the role of geography on scientific authorship of literary fiction and that of the academic logic on knowledge transfer. We advance some tentative conclusions regarding the incorporation of scientific authorship of literary fiction in the evaluation of academic merit.