Choosing a preprint server is a decision about five factors, not one: screening policy, licensing terms, persistent-identifier assignment, indexing reach, and journal-submission integration. A preprint server is an online repository that makes a research manuscript publicly available before formal peer review, typically assigning a DOI and an open licence so the work is citable immediately. Matching those five factors to a discipline, a funder mandate, and a target journal’s policy is what separates a defensible institutional recommendation from “just use bioRxiv.”
- What is a preprint server?
- Which five criteria actually differentiate preprint servers?
- How do the major preprint servers compare?
- Which preprint server fits your discipline?
- What should research offices check before recommending a server?
- Common questions about preprint servers
- What this means for researchers and institutions
What is a preprint server?
A preprint server is a repository, usually free to use, that publishes a manuscript version before it has completed formal peer review. Unlike a personal website or an institutional repository, a genuine preprint server performs basic screening, assigns a persistent identifier, and makes the work discoverable through indexing services and search engines.
The category now spans general-purpose infrastructure (arXiv, OSF Preprints), discipline-specific platforms (bioRxiv, medRxiv, ChemRxiv, PsyArXiv, SocArXiv), and publisher-operated services (SSRN, Research Square). Each applies a different mix of screening, licensing, and indexing — which is exactly why a single “best” answer does not exist.
Which five criteria actually differentiate preprint servers?
Most comparisons stop at subject scope. A more useful framework adds four operational criteria that determine whether a preprint actually functions as a citable, fundable, publishable output.
- Screening policy. Every reputable server checks for plagiarism, offensive or non-scientific content, and (for health research) potential public-harm risk. medRxiv applies the strictest clinical-harm review of the major platforms, reflecting the direct patient-facing risk of unreviewed medical claims.
- Licensing. Authors typically retain copyright but must grant a distribution licence. Options range from CC0 and CC BY through to CC BY-NC-ND or a non-exclusive posting licence with no reuse rights. cOAlition S’s Plan S rights-retention strategy specifically favours CC BY preprints, so funder compliance can hinge on this single field.
- Persistent-identifier assignment. A DOI — usually registered through DataCite or Crossref — is what makes a preprint permanently citable and trackable through Altmetric and citation indexes. Not every platform assigned DOIs from launch; arXiv began registering DOIs for new submissions via DataCite only in 2022, decades after its 1991 founding at Cornell University.
- Indexing and discoverability. Google Scholar indexes most major servers, but subject-specific indexing varies: medRxiv and bioRxiv preprints are only selectively surfaced in PubMed and PubMed Central, while SSRN content feeds RePEc for economics.
- Journal-submission integration. Direct-transfer tools — such as bioRxiv’s “B2J” (bioRxiv-to-journal) pipeline — let authors submit straight from the preprint record, cutting duplicate uploads and formatting work.
How do the major preprint servers compare?
The table below summarises governance, screening, and identifier practice across the platforms institutions most commonly encounter. Details change; always confirm current policy on the platform itself before advising a researcher.
| Server | Primary field | Governance | Screening | Licence options | PID / indexing |
|---|---|---|---|---|---|
| arXiv | Physics, maths, CS, quantitative biology | Cornell University Library (non-profit); founded 1991 | Moderator scope/format check | Non-exclusive licence or CC BY/CC BY-NC-SA | DataCite DOI (since 2022); Google Scholar, Semantic Scholar |
| bioRxiv | Biology, life sciences | openRxiv, an independent non-profit launched in 2024 to steward bioRxiv and medRxiv | Staff/advisor plagiarism and ethics check | CC BY, CC BY-NC, CC BY-ND, or no reuse | DOI on posting; Google Scholar, selective PubMed Central |
| medRxiv | Health and clinical sciences | openRxiv, with BMJ and Yale as founding partners; launched 2019 | Additional clinical-harm review beyond bioRxiv’s checks | CC BY, CC BY-NC, CC BY-ND, or no reuse | DOI on posting; explicit “not for clinical guidance” disclaimer |
| SSRN | Social sciences, economics, law | Elsevier (for-profit; acquired 2016) | Light editorial review | Author retains copyright under SSRN’s posting terms | Unique SSRN ID; feeds RePEc, Google Scholar |
| ChemRxiv | Chemistry | Operated with the American Chemical Society and partner chemistry societies | Technical and ethics moderation | CC BY, CC BY-NC, CC0 | DataCite DOI; indexed by CAS, Google Scholar |
| Research Square | Multidisciplinary; “In Review” journal integration | For-profit, partnered with Springer Nature | Basic format/ethics check | CC BY and other CC variants | DOI on posting; linked directly to journal submission systems |
| Preprints.org | Multidisciplinary | Operated by MDPI | Editorial screening, typically within days | CC BY by default | DOI via Crossref; Google Scholar indexed |
Which preprint server fits your discipline?
Discipline norms still drive the first cut of any decision. Physicists, mathematicians, and computer scientists default to arXiv because that is where the citation graph and community feedback already live. Life scientists post to bioRxiv, and clinical researchers use medRxiv precisely because of its stricter harm-review layer. Economists, legal scholars, and business researchers rely on SSRN’s established readership, while chemists increasingly use ChemRxiv because of its direct society backing.
Researchers working across disciplines, or without an obvious subject-specific home, should consider general-purpose infrastructure such as OSF Preprints or a multidisciplinary commercial platform such as Preprints.org — provided the licensing and screening terms still meet funder requirements.
What should research offices check before recommending a server?
Research-administration and library staff advising faculty need a repeatable checklist, not a single favourite platform:
- Does the target journal’s policy permit prior posting on this specific server? The ICMJE Recommendations state that posting a preprint does not constitute prior publication for most biomedical journals, but individual journal policies can still vary.
- Does the licence offered satisfy the researcher’s funder mandate — for example, UKRI’s or cOAlition S’s preference for CC BY on outputs arising from funded grants?
- Will the platform assign a DOI immediately, and is that DOI registered with DataCite or Crossref for downstream citation tracking?
- Is the server listed in a recognised directory — such as COAR’s Directory of Open Access Preprint Repositories or the ASAPbio preprint server list — that documents its screening and governance practices?
- Does the platform provide version control that clearly links a preprint to its eventual peer-reviewed publication?
COPE guidance reinforces that editors and authors should treat preprint disclosure transparently in submission and review workflows, which makes documented screening and licensing practice — not brand recognition — the correct basis for an institutional recommendation.
Common questions about preprint servers
What is a preprint server?
A preprint server is a repository that publishes a research manuscript before formal peer review, applying basic screening, assigning a persistent identifier such as a DOI, and making the work openly discoverable. It differs from a personal or institutional webpage by offering structured metadata, licensing, and indexing.
What are the disadvantages of preprints?
Unreviewed findings can be misreported by media before validation, negative public comments can occur before formal review, and posting adds an extra step to the publication timeline. In clinical fields, this is why medRxiv carries an explicit disclaimer against using preprints to guide clinical practice.
Which is the best preprint server?
There is no single best server — the right choice depends on discipline, target-journal policy, funder licensing requirements, and whether journal-submission integration matters. arXiv, bioRxiv, medRxiv, ChemRxiv, and SSRN each dominate a different subject area rather than competing directly.
How does a preprint differ from peer review?
A preprint has passed only basic screening for plagiarism and scope, while peer review involves independent experts formally assessing methodology, evidence, and conclusions. Peer review certifies a paper for a journal; a preprint server does not — it only makes a manuscript public and citable ahead of that certification.
What this means for researchers and institutions
Preprint infrastructure is consolidating around named, accountable stewardship rather than informal hosting. openRxiv’s 2024 launch as an independent non-profit overseeing bioRxiv and medRxiv is a governance signal research offices should track, alongside continuing publisher involvement through SSRN (Elsevier) and Research Square (Springer Nature). Neither model is inherently wrong, but each carries different long-term sustainability and independence trade-offs that belong in any institutional recommendation, not just in the author’s personal choice.
For research administrators, the practical output of this framework is a short internal guidance note: name the approved server per discipline, confirm its licence matches funder mandates, and confirm its DOI and indexing practice before telling researchers where to post. That single document does more to reduce compliance risk than any general preprint policy statement.
For related definitions and standards context, see the CASRAI Dictionary and guidance for research administration teams building institutional open-research policy.
