Finding and using preprints
If recency matters to your question, the preprint servers are not optional.
Where can I find preprints?
By field: arXiv for physics, mathematics and computer science, bioRxiv and medRxiv for life sciences and medicine, ChemRxiv for chemistry, SSRN for social sciences and law, and EarthArXiv, PsyArXiv and others for their disciplines. Europe PMC and OpenAlex index preprints alongside published work, so a single search there covers several servers at once.
Why preprints matter to a search
The gap between a result existing and appearing as a peer-reviewed paper is typically six to eighteen months, and often longer. A literature search that excludes preprints excludes the current state of any fast-moving field by construction.
They also change what is findable. Negative results, replications and methods papers are all posted as preprints at higher rates than they are eventually published, so the preprint literature is systematically less affected by publication bias than the journal literature is.
Where to look, by field
| Server | Fields | Notes |
|---|---|---|
| arXiv | Physics, maths, CS, quantitative biology, statistics, economics | The oldest and largest; moderated, not peer reviewed |
| bioRxiv | Biology | Screened before posting; links to the published version when one appears |
| medRxiv | Health and clinical research | Additional screening; carries an explicit clinical-use caution |
| ChemRxiv | Chemistry | Run by a consortium of chemical societies |
| SSRN | Social sciences, law, economics, humanities | Owned by Elsevier; working papers as well as preprints |
| Research Square | Multidisciplinary | Often the in-review version of a submitted manuscript |
| PsyArXiv, EarthArXiv, SocArXiv, EngrXiv | Psychology, earth sciences, sociology, engineering | Community-run on the Open Science Framework |
| OSF Preprints | Multidisciplinary aggregator | Searches many of the above at once |
For a single search across several servers plus the published literature, Europe PMC, OpenAlex and Semantic Scholar all index preprints. Google Scholar does too, but does not let you filter to them.
Always check for a published version
A large share of preprints are eventually published, and the published version can differ substantially — analyses change, claims are qualified, and occasionally a central result does not survive review.
- Check the preprint landing page. bioRxiv, medRxiv and most modern servers display a link to the published article when one is registered.
- Search the exact title in Crossref, OpenAlex or Google Scholar. A published version will usually be the first hit.
- Check the preprint DOI in Crossref, which records the relation between preprint and version of record.
- If a published version exists, read and cite that one. Note any difference from the preprint if you had already relied on it.
How to cite a preprint
Cite the preprint DOI, name the server, give the posting date, and label it as a preprint. Most style guides now have an explicit form for this, and the label is the part that matters: a reader must be able to tell from the citation that the work has not been peer reviewed.
Include the version number where the server assigns one. Preprints are revised in place, and a citation without a version can point at text that no longer exists.
Some journals restrict citing preprints, particularly in clinical fields. Check the target journal instructions before building an argument on one.
Reading a preprint carefully
- Screening is not review. Servers check for scope, plagiarism and obvious problems; nobody has assessed the methods.
- Check whether the data and code are available. A preprint with an open dataset is more checkable than a reviewed paper without one.
- Look at the comment threads. Several servers host public comments, and substantive methodological objections often appear there long before any published critique.
- Treat a single unreplicated preprint result as a hypothesis. That is not scepticism about preprints specifically — it is the same standard a single unreplicated published result deserves.
Questions
- Are preprints peer reviewed?
- No. Servers screen submissions for scope, ethics and obvious problems, but no reviewer has assessed the methods or the conclusions. Some preprints later carry post-publication review or community comments; that is not the same as journal peer review.
- Can I cite a preprint in a thesis or paper?
- Generally yes, if you label it as a preprint, give the DOI and the version, and do not lean on it for a claim that needs settled evidence. Some journals and some clinical fields restrict it, so check the instructions for authors first.
- How do I know if a preprint was later published?
- Check the landing page, which usually links to the published version, and search the exact title in Crossref or OpenAlex. Crossref records the preprint-to-version-of-record relation, so the DOI itself will often resolve the question.
- Do preprints count for systematic reviews?
- Increasingly yes, as a source of unpublished evidence that reduces publication bias, and several guidelines now address how to include them. Report them separately from peer-reviewed studies and take their unreviewed status into account when assessing certainty.
Related guides
- How to find research papers on a topic — How do I find research papers on a topic?
- Searching grey literature — What is grey literature and how do you search it?
- Getting papers you cannot access — How can I legally access a paywalled research paper?
- Google Scholar alternatives — What are the best alternatives to Google Scholar?
Last updated 2026-08-25. Part of the litscout literature search guides.