An author list rarely stays frozen from the moment a manuscript first goes up as a preprint to the day the peer-reviewed version of record appears. Contributors join after revision experiments, a co-author withdraws, order gets renegotiated once reviewers weigh in on who did what — and each of these has a different mechanical consequence depending on which preprint server hosted the original posting. This guide covers how often author lists actually change between preprint and published version, why it happens, and what each of the major life-science preprint platforms (bioRxiv, medRxiv, and Preprints.org) requires when it does.
For the mechanics of preprint version numbers themselves — how v1/v2/v3 works and how citations should point to a specific version — see Preprint Version Numbers (v1, v2, v3). For changing author order specifically while a manuscript is still in draft, see How to Change Author Order Mid-Draft. For what happens when a publisher (not a preprint server) needs to correct a name after formal publication, see Retroactive Author Name Change Policy. This page sits between those two: the specific window where a manuscript already exists as a public preprint but hasn’t yet become the version of record.
How often does the author list actually change?
The best available empirical evidence comes from a 2022 PLOS Biology study, Brierley et al., “Tracking changes between preprint posting and journal publication during a pandemic” (PLOS Biology 20(2): e3001285), which compared preprints against their eventual published counterparts across a large sample spanning bioRxiv and medRxiv. Its topline finding: the large majority of preprints — over 85% — showed no authorship changes at all between the preprint and the published version. Most manuscripts that go up as a preprint come out the other end of peer review with exactly the same author list they started with.
Where changes did happen, adding authors was far more common than removing them. The study also found a notable COVID-19 effect: preprints on COVID-19 topics were roughly three times as likely to gain additional authors on publication as non-COVID-19 preprints (about 17.2% versus 6.2%), plausibly reflecting the compressed, collaborative, fast-moving nature of pandemic-era research, where a project could accumulate new contributors (additional data, additional sites, additional analysis) between an early preprint posting and a later, more complete published version. Within the bioRxiv-specific portion of the sample, the study found no instances of author removals or corresponding-author changes at all — consistent with addition being the dominant type of change, when a change happens at all.
None of this means an author-list change is a red flag. It is a normal, expected consequence of what a preprint actually is: a snapshot of a manuscript at one point in an ongoing research and revision process, not a final, fixed record. New analyses get added during peer review, a co-author who contributed a key revision experiment gets added, or someone originally listed steps back. What is unusual is silent, undisclosed reordering with no paper trail — that is a matter for ICMJE-aligned authorship criteria and, if it becomes contentious, COPE-style dispute handling, not for the preprint server itself.
What actually triggers a required new preprint version
This is the part that genuinely differs by platform, and treating it as one uniform rule is the most common mistake researchers make. Verify against each platform’s own current author guidance before assuming a mechanism carries over from one server to another.
bioRxiv and medRxiv (openRxiv)
bioRxiv and medRxiv have shared submission and screening infrastructure since both moved under the nonprofit openRxiv in March 2025 (see How to Submit to bioRxiv or medRxiv for the full submission mechanics), and their author-list policies track together. Per their published FAQ guidance, the servers draw a specific, narrow distinction:
- “Silent” name changes — an update to one individual author’s own first or last name (for example, following a legal name change) — can be made without a new version or a correction notice, by emailing the editorial office directly. No explanation is required for this category.
- Everything else — adding an author, removing an author, or reordering the byline — is not covered by that silent-change provision. The stated mechanism is to post a revised version of the preprint. openRxiv’s own guidance notes that most authors prefer this route anyway, since a new version is what actually propagates the change to third-party indexers and citation databases, where a quiet email-only edit might not.
Because bioRxiv/medRxiv keep every version permanently accessible under the same DOI (see Preprint Version Numbers for how that works), posting a new version to reflect an author-list change doesn’t erase the record of who was listed before — both the old and new author lists remain visible in the preprint’s version history, which is a meaningful difference from a silent, unversioned edit.
Preprints.org (MDPI)
Preprints.org runs its own versioning system, independent of the ICMJE-adjacent openRxiv model above. Substantive authorship changes — adding or removing an author — are generally handled through documentation similar to a journal-style authorship change process: identifying what changed, briefly explaining why, and confirming agreement from the affected authors, broadly consistent with the authorship-change-form pattern used across MDPI’s own journals and by most major publishers (Elsevier, ACS, Frontiers, and others all publish a comparable form for post-submission authorship changes). Minor corrections — a misspelled name, a wrong affiliation, byline order — are typically resolved by contacting the platform directly rather than by posting a full new version.
A caveat worth being explicit about: Preprints.org’s own detailed versioning and author-change policy pages returned an access error during verification for this guide, so treat the specifics above as a lower-confidence, general characterization rather than a verbatim platform rule. If a change to authorship on a Preprints.org submission is time-sensitive, confirm the current procedure directly against Preprints.org’s own author instructions or by contacting their editorial office before assuming it matches the bioRxiv/medRxiv model.
The pattern across platforms, and why it isn’t uniform
The common thread is that every platform distinguishes between cosmetic/individual corrections (a misspelled name, an affiliation typo, a legal name update) — which can usually be handled without a new version — and substantive changes to who is credited as an author at all, which generally do require a new version or a documented, author-agreed change process. What differs is where each platform draws that line and how much paperwork sits behind it. Don’t assume a Preprints.org process maps onto bioRxiv’s, or that either maps onto how a publisher handles a post-publication correction (see Retroactive Author Name Change Policy for that separate, later-stage process).
Practical guidance for researchers
- Settle authorship before the first posting where possible. Applying ICMJE’s four authorship criteria at the point a manuscript first goes up as a preprint avoids a version cascade later. This is not always possible — a fast-moving or multi-site project may genuinely add a qualifying contributor after the first posting — but it is the cheapest way to avoid the problem entirely.
- Get agreement in writing before requesting a change. Every platform’s mechanism assumes the current author list agrees to the change. Get that agreement, ideally in an email thread the corresponding author can point to, before submitting a revision request — this is exactly the kind of dispute COPE’s authorship-change guidance exists to prevent from escalating.
- Expect the old author list to stay visible. On bioRxiv/medRxiv specifically, prior versions remain publicly accessible under the same DOI. A new version doesn’t erase the earlier byline — it adds a dated record of the change. Don’t treat a version bump as a way to quietly overwrite history.
- Check whether your citation needs updating. If you cited an earlier preprint version by its version-specific URL and the author list has since changed in a later version, verify you’re pointing at the version you actually intended to cite — see Preprint Version Numbers for the mechanics of version-specific citation.
- Don’t assume the published version will match the preprint’s final author list either. Journal peer review can prompt its own authorship changes — a new co-author added because of a reviewer-requested analysis, for instance — independent of anything that happened at the preprint stage. The preprint’s last version and the published version of record are two separate checkpoints, each with its own change history.
Frequently asked questions
Is it normal for authors to be added between preprint posting and journal publication?
Yes, though it is the exception rather than the rule. Empirical data (Brierley et al., 2022, PLOS Biology) found over 85% of preprints carry an unchanged author list through to publication; where changes occur, adding authors is more common than removing them, and the effect is notably stronger for fast-moving, highly collaborative research areas.
Does changing the author list on a preprint delete the old version?
No, at least not on bioRxiv or medRxiv. Both platforms retain every prior version under the same DOI, so the earlier author list remains publicly visible in the preprint’s version history even after a new version reflects the updated byline.
Can I just email the preprint server to change an author’s name?
On bioRxiv and medRxiv, that “silent” email-only route is available specifically for an individual author’s own name update (for example, a legal name change) — it is not available for adding, removing, or reordering authors, which those platforms handle by posting a new version instead.
Does a preprint author-list change need to match what eventually gets published?
Not necessarily. The preprint’s author list at its last version and the journal’s published version of record are governed by separate processes — a journal’s peer-review process can introduce its own authorship changes independent of anything already resolved at the preprint stage.
