Scientific literature is the totality of publications that report original empirical and theoretical work in the sciences and social sciences. It is often abbreviated to the literature. Academic publishing is the process of placing one's research work into the literature.

The scientific literature includes the following kinds of publication:

• almost all articles published in scientific journals
• patents
• some "authored books", that is books wholly written by one or a small number of authors
• chapters in many "edited books", that is, books where each chapter is the responsibility of a different author or set of authors, though the editor or editors may take some responsibility for ensuring consistency of style and content
• some presentations at academic conferences , especially those organized by learned societies
• some government reports
• some publications on the World Wide Web
• some books, pamphlets and reports issued by individual researchers or research organisations on their own initiative; sometimes these are organised into series of "Working Papers" or "Technical Reports".

The importance of these components of the literature varies between disciplines and has varied over time. At present (2005) it is true in almost all disciplines that the standing of journal article publication is the highest (though journals vary very greatly in their prestige and importance) and that of Working Paper series and World Wide Web publication is the lowest (though with some important exceptions). The standing of book publications is highly variable.

Ultimately, it is not the format that is important, but what lies behind it. Several key requirements have to be met before an outlet can be regarded as forming part of the literature.

1. There should have been some element of peer review of the content. The lack of peer review is what makes most Technical Reports and World Wide Web publications unacceptable as contributions to the literature. The relatively weak peer review often applied to books and chapters in edited books means that their status is doubtful, unless an author's personal standing is so high that his or her prior career provides an effective guarantee of quality.
2. The format should be archival, in the sense that libraries should be able to store and catalogue the documents and scientists years later should be able to recover any document in order to study and assess it, and there should be an established way of citing the document so that formal reference can be made to them in future scientific publication. The lack of an established archival system is one of the hurdles that World Wide Web based scientific publication has to overcome; significant progress is now being made on this front.
3. The content should be located in the context of previous scientific investigations, by citation of relevant documents in the existing literature.
4. Empirical techniques, and the results of the investigation, should be described in such a way that a subsequent scientist, with appropriate knowledge of and experience in the relevant field, should be able to repeat the observations and know whether he or she has obtained the same result.
5. The conclusions drawn should be based on previous literature and/or new empirical results, in such a way that any reader with knowledge of the field can follow the argument and confirm that the conclusions are sound. That is, acceptance of the conclusions must not depend on personal authority, rhetorical skill, or faith.

Peer review and the learned journal format are convenient ways of ensuring that these fundamental criteria are met, rather than being in themselves essential to a scientific literature.

Increasing reliance on abstracting services , especially on those available electronically, means that the effective criterion for whether a publication format forms part of the literature is whether it is covered by these services; in particular, by the specialised service for the discipline concerned such as Chemical Abstracts Service, and by the major interdisciplinary services such as those marketed by the Institute for Scientific Information.