Preprint vs Published Paper: What Is the Difference, and Can You Cite a Preprint?
How a preprint differs from an accepted manuscript and a published paper, how much preprints change before publication, and when and how to cite one.
When you find a free copy of a paper, it is often labelled a preprint or an accepted manuscript. These are real versions of the paper, shared legally by the authors, and the difference between them matters for how much you can rely on the text.
The three versions of a paper
- Preprint. The manuscript as the authors wrote it, posted before peer review on a server such as arXiv, bioRxiv, medRxiv or SSRN. It is also called the submitted version.
- Accepted manuscript. The text after peer review and revision, as the journal accepted it, but before typesetting. It is sometimes called a postprint.
- Published version. The final article in the journal's layout, with page numbers. It is also called the version of record.
Our finder labels each free copy with one of these, so you know which you are opening.
How different is a preprint from the published paper?
Usually not very. In most cases the main findings and conclusions are the same, and the changes made during review are clarifications, extra analyses and corrections. Sometimes the changes are larger: a conclusion is softened, an error is fixed or an experiment is added. Occasionally a preprint is never published at all, which can mean it did not survive review.
The accepted manuscript is a different matter. Its content is the same as the published paper. Only the formatting differs.
Is it safe to rely on a preprint?
For understanding what a study did and found, a preprint is generally fine. Be more careful when the stakes are high, for example with medical claims. Check three things:
- Has it since been published? Preprint servers add a link to the journal version once it exists. If it has, read or cite that.
- How old is it? A preprint that is several years old and never published deserves some caution.
- Which version are you reading? Authors can update a preprint. The latest version is usually the most accurate.
Our guide to telling whether an article is peer reviewed covers the checks in more detail.
Can you cite a preprint?
Yes. Preprints are citable, and every major citation style has a format for them. Two rules of thumb apply:
- If a published version exists, cite that one, even if the preprint is what you read. The published version is the permanent record.
- If only the preprint exists, cite it and say that it is a preprint. Give the authors, the year, the title, the name of the server and the DOI or link.
Some teachers and journals limit how many preprints you may cite, so check any guidance you have been given. Our guide to citing a research paper shows the main formats.
What if you read the accepted manuscript?
Cite the published version. The content is the same, and the published version is the one readers can look up. The only practical difficulty is page numbers: if you need to quote a specific page, the published layout is the one to refer to. If you cannot see it, cite the article as a whole.
Finding a preprint or accepted manuscript
Paste a paper's DOI, link or title into the finder. If a preprint or accepted manuscript exists in a repository, it will be listed. Our guide to arXiv explains the largest preprint server.
Find a free copy of a paper
Paste a DOI, a link to the paper or its title, and we will look for a free, legal copy. No signup.
Frequently asked questions
Is a preprint a real publication?
A preprint is a public, citable version of a paper, but it has not been peer reviewed and does not count as publication in a journal.
What is the difference between a preprint and a postprint?
A preprint is the version before peer review. A postprint, more often called the accepted manuscript, is the version after peer review but before the journal’s typesetting.
Should I cite the preprint or the published paper?
Cite the published paper whenever one exists. Cite the preprint only when the work has not yet been published, and state that it is a preprint.