Preprint Server Policies: Avoiding a “Prior Publication” Conflict Before Journal Submission

The short answer

Posting a manuscript on a recognized preprint server before journal submission is not considered “prior publication” by any major biomedical or scientific publisher today. This is the settled, publisher-confirmed answer as of 2026: the era when posting a preprint could disqualify a manuscript under the so-called “Ingelfinger rule” is over for essentially every major publisher. But “not disqualifying” is not the same as “no rules apply.” Journals still expect disclosure, still expect the preprint to be updated to point to the published version, and a small number still apply narrow, publisher-specific restrictions (media handling, specific server choice, or field-specific caveats). This guide walks through what actually counts as a conflict, what the major publishers say in their own policies, and the concrete steps to check before you post or submit.

What “prior publication” actually means

“Prior publication” (sometimes called duplicate or redundant publication) is the rule against submitting the same work, in substantially the same form, to more than one venue without disclosure — either as a formally published article elsewhere, or simultaneously under review at more than one journal. The International Committee of Medical Journal Editors (ICMJE) states the underlying rule plainly: “Authors should not submit the same manuscript, in the same or different languages, simultaneously to more than one journal,” and authors who attempt this “should expect at least prompt rejection of the submitted manuscript” — or retraction if it is discovered after publication.

That rule was historically read very broadly. The colloquial name for the strict version — the “Ingelfinger rule,” after a former New England Journal of Medicine editor who in 1969 announced NEJM would not consider work that had already been reported elsewhere, including in the press — treated almost any public disclosure of results before formal peer-reviewed publication as disqualifying. Preprints did not meaningfully exist as an author-facing practice in most fields when that norm took hold, so for decades the safest assumption for an author was: don’t post or present findings publicly before your target journal has accepted them.

Why preprints are treated differently now

That assumption is now out of date for the great majority of journals and fields. ICMJE’s own Recommendations directly address preprints and state that posting on a preprint server does not, by itself, constitute prior or duplicate publication. ICMJE does add conditions:

• Authors should choose a preprint server that clearly labels postings as not peer reviewed, discloses authors’ conflicts of interest and funding sources, retains metadata for withdrawn preprints, and gives authors a way to link the preprint to the eventual published version.
• Authors must inform the journal at submission that the work was posted as a preprint.
• Authors are responsible for updating the preprint to point readers to the peer-reviewed version once it is published — not the journal.
• Posting as a preprint “may influence a journal’s interest in or priority for peer review and publication of that work” — ICMJE’s own wording, meaning some editors may weigh it as a factor even though it is not a bar.
• The final, peer-reviewed article should not be re-posted to the preprint archive in place of the original preprint.

In practice, this means the conflict to avoid is not “did I post a preprint,” it’s “did I fail to disclose it, fail to update it, or violate a specific journal’s narrower rule on top of the general allowance.”

What the major publishers actually say

General allowance is the norm, but publisher-level policy pages are where the real, checkable specifics live. A research administrator advising an author should point them to the actual policy page for the target journal’s publisher, not rely on a general impression. Several worth knowing by name:

Springer Nature

Springer Nature has a unified, cross-journal preprint policy: “Posting of preprints of research manuscripts is not considered prior publication and will not jeopardize consideration at Springer Nature journals.” Authors may post to “any legal preprint sharing platform of your choice,” and Springer Nature supports any variety of preprint license, including Creative Commons licenses. Springer Nature also operates its own journal-integrated preprint mechanism, In Review (built with Research Square), which posts a preprint automatically at the point of submission for participating journals, for authors who opt in. One genuinely specific caveat Springer Nature flags: authors who discuss a preprint with the media should make clear that the findings are not yet peer reviewed and may change before final publication — a media-handling note, not a publication-status restriction.

Elsevier

Elsevier’s sharing policy likewise treats preprint posting as compatible with subsequent journal submission, and several Elsevier journals offer a preprint-posting option (via SSRN) directly inside the submission workflow, so authors can post concurrently with submitting rather than as two separate steps. As with any publisher, the exact wording and any subject-specific caveats live on the individual journal’s “Guide for Authors” page — Elsevier’s own preprint policy documentation has been noted, including by independent analyses of publisher preprint-policy clarity, as less consistently worded across its journal portfolio than Springer Nature’s unified policy, which is itself a practical reason to check the specific target journal rather than assuming a blanket Elsevier-wide rule.

PLOS

PLOS journals actively encourage preprinting rather than merely tolerating it. PLOS supports direct submission from bioRxiv and medRxiv — authors post a preprint on either server and can concurrently submit the same manuscript to a PLOS journal through a direct-transfer integration, without re-uploading files. PLOS Biology also offers a facilitated posting option: authors can opt in to have PLOS forward a submitted manuscript to bioRxiv for a preprint posting after an initial suitability screen. Individual PLOS journals (PLOS Medicine, PLOS Biology, PLOS Computational Biology, and others) each publish their own preprints page under their author guidelines, which is the place to confirm current specifics for that particular journal.

Nature-family and Science-family journals

Nature-family journals (Springer Nature) fall under the unified Springer Nature policy above. Science-family journals (published by AAAS) and other major biomedical and life-science titles (Cell, NEJM, JAMA, The Lancet) now generally permit preprint deposit ahead of submission as well, reflecting the same broad industry shift — but AAAS’s policy has historically been more conservative about press coverage of preprinted findings than some other publishers, and the exact current wording should be checked directly on the journal’s own author-guidelines page before relying on it, since this is exactly the kind of publisher-specific detail that changes without much notice and where a general summary can go stale.

Which preprint servers are broadly accepted

None of the major publisher policies above restrict authors to a single named server; ICMJE’s criteria (peer-review labeling, disclosure requirements, withdrawal metadata, a published-version link mechanism) function as the practical bar a server needs to clear, not a specific list. In practice, the servers that meet this bar and are broadly recognized by publishers and funders include:

• arXiv — physics, mathematics, computer science, quantitative biology/finance, statistics.
• bioRxiv and medRxiv — life sciences and clinical/health sciences respectively, jointly operated by openRxiv.
• OSF Preprints — a multi-discipline aggregator built on the Open Science Framework.
• SSRN — social science, economics, law, and business.
• Research Square — general-discipline, and the engine behind Springer Nature’s In Review.
• Field-specific servers such as ChemRxiv (chemistry) and PsyArXiv (psychology).

For side-by-side detail on how two of these differ operationally, see CASRAI’s comparisons of arXiv vs. SSRN and bioRxiv vs. medRxiv, and the practical walkthrough for submitting to bioRxiv or medRxiv. For the criteria to weigh when a server isn’t already dictated by your field or funder, see How to Choose a Preprint Server.

Where conflicts genuinely still happen

Given how broad the general allowance now is, the real risk for most authors isn’t posting a preprint at all — it’s one of these narrower failure modes:

• Not disclosing the preprint at submission. ICMJE and most major publishers require authors to state, at submission, that a preprint exists. Omitting this is the disclosure failure that turns an otherwise-fine preprint into a duplicate-publication problem if an editor discovers it independently.
• Submitting to a journal with a genuinely stricter, non-default policy. A small number of journals — often in fields or subfields with different sharing norms, or specific society-run titles with their own editorial policies — still restrict or prohibit preprints, or restrict them for certain article types (e.g., invited reviews, case reports) even when they allow them for original research. This is exactly why “most publishers allow it” cannot substitute for checking the specific target journal.
• Reposting the final, formatted, peer-reviewed article back to the preprint server instead of just linking to it — this can conflict with the journal’s copyright/licensing terms even when the original preprint posting was fine.
• Simultaneous submission to two journals — a preprint posting is not simultaneous submission (it isn’t a competing journal), but posting the same manuscript as a preprint at two different preprint servers, or actually submitting the manuscript to two journals in parallel, is the classic overlapping-publication problem ICMJE describes, independent of preprints.
• Confusing preprints with conference proceedings or theses/dissertations. These raise related but distinct “prior dissemination” questions with their own publisher-specific answers — don’t assume a preprint-friendly journal automatically treats a conference paper or dissertation chapter the same way.

A practical pre-submission checklist

1. Read the specific target journal’s Instructions for Authors / Guide for Authors, not just the publisher’s general policy page. ICMJE explicitly recommends journals state their preprint policy in their author instructions — use that as the authoritative source, in preference to a general publisher-wide summary (including this one).
2. Confirm the preprint server you intend to use (or already used) meets basic disclosure standards — clear “not peer reviewed” labeling, conflict-of-interest and funding disclosure, and a mechanism to link to the eventual published version. Most of the servers listed above already meet this bar.
3. Disclose the preprint at submission, typically in the cover letter and/or a dedicated submission-system field — see CASRAI’s guide on conflict-of-interest disclosure in scholarly publication for related disclosure practice.
4. Check whether the article type matters. Some journals that allow preprints for original research restrict the practice for other article types (reviews, commentaries, case reports) — confirm this applies (or doesn’t) to what you’re submitting.
5. Plan to update the preprint once the paper is accepted — add the DOI and citation for the published version. ICMJE places this responsibility on authors, not the journal or the preprint server.
6. If in doubt, ask the editorial office directly before posting, particularly for journals in fields where preprinting is newer or less standard, or for co-authored, multi-institutional, or funder-mandated work where more than one policy might apply at once.

Frequently asked questions

Does posting a preprint count as “prior publication”?

Not under the policies of the major publishers covered above, and not under ICMJE’s Recommendations, provided the preprint is disclosed at submission and posted to a server that meets basic transparency standards (non-peer-review labeling, disclosure, a link to the eventual published version). It can still affect an individual journal’s editorial priority, and a minority of journals apply stricter, narrower rules — always confirm against the specific target journal.

Do I need to tell the journal I posted a preprint?

Yes. ICMJE and the major publisher policies place this responsibility explicitly on the author, typically disclosed in the cover letter or a dedicated field in the submission system.

Can I submit to more than one journal at the same time if one copy is “just” a preprint?

No — a preprint posting is not the same thing as simultaneous journal submission, but submitting the identical manuscript to two journals at once (preprint or not) is the overlapping-publication problem ICMJE describes, and remains grounds for rejection or retraction.

What is the Ingelfinger rule, and does it still apply?

It’s the historical norm, associated with a 1969 NEJM editorial policy, of not considering work that had already been publicly disclosed elsewhere. In its strict original form it no longer applies at the publishers covered in this guide with respect to preprints specifically — all now treat preprint posting as compatible with subsequent submission, subject to disclosure.

Does having a DOI on my preprint change anything?

A preprint DOI makes the work citable and easier to link back to from the eventual published article, which supports the “point readers to the published version” step ICMJE asks authors to handle — but it does not, by itself, change whether the posting counts as prior publication either way.

Are all preprint servers treated the same by journals?

Not automatically. Publisher policies generally describe criteria a server should meet (peer-review labeling, disclosures, withdrawal handling) rather than naming one required server, and the servers listed in this guide are broadly recognized as meeting that bar. A journal that has never encountered a less common or newer server may ask questions before proceeding — check with the editorial office if your intended server isn’t a well-established one.