A related guide on this site, Preprint Server Policies: Avoiding a “Prior Publication” Conflict Before Journal Submission, addresses the author’s side of this question: whether and how to post a preprint before submitting to a journal. This guide addresses the other side of the desk — what a journal editor or editorial office should do when a submitted manuscript has already appeared, in some form, on a preprint server. The Committee on Publication Ethics (COPE) maintains a specific position document for this exact scenario, and its guidance is the primary reference this page works from.
COPE’s position, in brief
COPE’s position statement, Handling material previously published on a preprint server, sets out how editors should treat manuscripts with a preprint history. The short version: COPE does not treat posting to a recognized preprint server, on its own, as prior publication in the sense that would disqualify a manuscript from consideration. Editors are expected to have a clear, published policy on preprints rather than deciding case by case, and that policy should be visible on the journal’s website and built into the submission system so authors see it before they submit.
This builds on COPE’s broader definition of prior publication, which it describes as the publication, or attempted publication, of whole or substantial parts of a work, its data, or its analysis that have already been published or submitted elsewhere without transparency or appropriate declaration. The operative word is transparency: a disclosed preprint is a different situation, ethically, from an undisclosed duplicate submission. COPE draws the same distinction elsewhere — its companion position document, Handling material previously published as conference proceedings, uses the same logic for conference papers that later become journal submissions: the question is not whether the material appeared publicly before, but whether it was disclosed, and how much has substantively changed.
Why preprints are treated differently from other prior publication
Two things distinguish a preprint from the kind of prior publication editors are actually trying to screen out:
- It hasn’t been through peer review. A preprint is an author’s manuscript posted before or during the journal review process, not a competing version-of-record. See CASRAI’s preprint definition and, for contrast, version of record.
- It’s disclosed by design. Preprints are public, dated, and usually carry a DOI, which makes the prior version fully traceable — the opposite of the concealment problem prior-publication policies exist to catch.
This is also the framing in COPE’s 2018 discussion document, Best practices for preprints, written by COPE Council to open discussion rather than to hand editors a finished rulebook. It lays out both sides for editorial offices to weigh: preprints can accelerate communication of results, establish precedence, and give editors an early look at work that may later be submitted or transferred to their journal; they also raise real editorial questions around the risk of being “scooped,” how much screening a preprint server actually applies before posting, and how credible non-peer-reviewed material should be treated when it circulates ahead of formal publication. COPE frames these as open questions that vary by discipline and journal, not a single fixed rule — which is precisely why COPE’s advice to editors is to adopt and publish a clear policy rather than assume a house norm.
When a preprint history should make an editor look more closely
COPE’s guidance is not a blanket pass. A few situations genuinely warrant closer editorial scrutiny rather than an automatic assumption that “preprint means fine”:
- Substantial change between preprint and submission with no disclosure of what changed. If the submitted manuscript differs materially from the posted preprint — different results, added or removed authors, a different conclusion — the editor should understand why, the same way they would for any manuscript with an unexplained discrepancy from a related public document.
- Simultaneous or undisclosed submission to more than one journal. A preprint being public doesn’t change COPE’s normal expectation that a manuscript isn’t under review at more than one journal at once; that’s a separate rule from the preprint question and still applies.
- Disciplines or article types with specific expectations. COPE notes that prepublication posting requirements vary by field — clinical trial reporting is the clearest example, where prospective registration is expected regardless of preprint status, and some journals have their own supplementary requirements for trial manuscripts.
- Anonymity of peer review. Because a preprint discloses author identity publicly, a manuscript with a preprint history cannot realistically go through single- or double-anonymized review in the way a never-disclosed submission could — editors should factor this into which review model is workable and should not represent review as anonymized when the authors’ identities are already public.
Building this into the editorial workflow
Practically, COPE’s position translates into a small number of concrete steps for an editorial office:
- Publish the journal’s preprint policy somewhere authors will actually see it before submitting — the journal’s author guidelines and the submission system itself, not just a general ethics page.
- Ask directly, at submission, whether the manuscript (or a substantially similar version) has been posted as a preprint, and where. Making this a required submission field is more reliable than expecting voluntary disclosure in a cover letter.
- Record the preprint’s server, posting date, and DOI (if any) alongside the manuscript record, so the editorial office has a clear timeline if a prior-publication question comes up later.
- Decide, per the journal’s published policy, whether the preprint needs to be updated once the peer-reviewed version publishes. Many journals expect or require authors to link the preprint to the published article and note its status (e.g., that a peer-reviewed version now exists) — this is a version-control step editors can reasonably ask authors to complete rather than something the editorial office does unilaterally on a server it doesn’t control.
- Apply the same screening consistently, rather than case by case, so authors of similar manuscripts are treated the same way regardless of which editor handles the submission.
Version control between the preprint and the published article
Once a manuscript with a preprint history is accepted, editors and authors are dealing with two public copies of related work that need to be legible to readers, indexers, and citation databases as connected but distinct: the preprint (an author manuscript, not peer-reviewed) and the eventual version of record (the peer-reviewed, published article). Good practice, consistent with what preprint servers and COPE both point toward, is for the preprint listing to be updated after formal publication to note that a peer-reviewed version exists and to link to it — this is standard practice on servers such as arXiv and SSRN, and it is the mechanism that lets a reader who lands on the older preprint find their way to the current, peer-reviewed version rather than treating an outdated draft as final.
Related COPE and publisher resources
Editors handling this question in practice generally aren’t relying on a single document. COPE’s preprint position sits alongside a few other primary references worth having on hand:
- ICMJE’s recommendations on overlapping publications cover the closely related question of what counts as duplicate or redundant publication for medical and health-science journals, and are widely cross-referenced alongside COPE guidance by biomedical editorial offices.
- ASAPbio, a scientist-driven initiative focused on preprints in the life sciences, publishes best-practice recommendations on how preprints should be posted, screened, versioned, and linked to their published form — guidance several publishers now build directly into their own preprint policies.
- Individual publisher policies (Springer Nature, Elsevier, PLOS, and others) translate this general position into journal-specific submission instructions, since COPE’s own documents are deliberately framed as guidance for editors to adapt, not a single mandatory rule.
For the practical, discipline-by-discipline breakdown of which major preprint servers and publishers currently accept prior preprint posting, and where friction still occurs in practice, see the companion guide: Preprint Server Policies: Avoiding a “Prior Publication” Conflict Before Journal Submission.
Frequently asked questions
Does COPE consider a preprint to be a prior publication?
No, not on its own. COPE’s position is that posting to a preprint server does not, by itself, constitute the kind of prior publication its broader guidance is concerned with, because the definition of prior publication turns on transparency and disclosure, not on whether a version of the text has previously appeared anywhere in public. Editors are still expected to set and publish their own journal-specific policy rather than assume this applies uniformly.
Should editors require authors to disclose a preprint at submission?
Yes. COPE’s guidance assumes editors will ask, and building a preprint-disclosure field into the submission system is the reliable way to get that information consistently, rather than relying on authors to volunteer it.
Can a manuscript with a preprint history go through anonymized peer review?
Not in the way a manuscript with no public history can. Because the preprint discloses author identity, single- or double-anonymized review is not fully achievable once a reviewer can find the preprint; editors should choose a review model that reflects this rather than presenting the review as anonymized when it functionally isn’t.
Who updates the preprint once the peer-reviewed article is published?
This is generally the author’s responsibility, done through the preprint server’s own linking mechanism, though a journal’s policy can require it as a condition of publication. The editorial office does not typically have the ability to edit a listing on a server it doesn’t operate.
Does this guidance apply the same way in every discipline?
No. COPE explicitly notes that norms around prepublication posting vary by field, and some article types — clinical trial reports being the clearest case — carry their own prospective registration and disclosure expectations that apply regardless of preprint status.
