Every bioRxiv and medRxiv preprint receives a single, permanent DOI that stays constant across all revisions — the DOI always resolves to the newest version, and a specific version (v1, v2, v3) is cited by appending the version number to the DOI-based URL, not by requesting a new identifier.
A bioRxiv preprint DOI is a Crossref-registered digital object identifier assigned to a manuscript once openRxiv’s screening team approves it for posting, and it serves as the manuscript’s permanent citation handle for the life of the record. Understanding how that identifier behaves across revisions — and how it eventually connects to a journal’s version of record — is essential for anyone citing, tracking, or administering preprint outputs.
- What is a bioRxiv preprint DOI?
- How do bioRxiv and medRxiv assign DOIs across versions?
- How do you cite bioRxiv v1 vs v3 correctly?
- How does the preprint DOI link to the published journal article?
- Common questions about bioRxiv and medRxiv DOIs
- Why this matters for research administrators
What is a bioRxiv preprint DOI?
A bioRxiv or medRxiv DOI is issued the moment a submission clears openRxiv’s screening process, which bioRxiv’s Submission Guide states typically takes 24–72 hours. The identifier is deposited with Crossref, the DOI registration agency used by both servers, and it is what makes a preprint “citable and part of the scientific record” rather than a private working draft.
Since December 1, 2025, all newly posted bioRxiv and medRxiv articles use the prefix 10.64898, replacing the legacy 10.1101 prefix used throughout the Cold Spring Harbor Laboratory (CSHL) era. According to openRxiv’s own explainer, published November 18, 2025, articles posted before that date keep their existing 10.1101 DOIs unchanged — nothing needs to be re-cited or updated by authors or readers; only new submissions carry the new prefix.
The DOI suffix is not arbitrary. Since December 11, 2019, it has embedded the date the author approved the submission for posting (e.g., 2026.01.01.123456), which lets a reader estimate an article’s age directly from the citation string, much as a volume and year do for a journal article. DOIs assigned before that date used a simple six-digit suffix instead.
How do bioRxiv and medRxiv assign DOIs across versions?
openRxiv assigns exactly one DOI per article, and that single DOI covers every subsequent version. When an author submits a revision, the new version posts under the same DOI; the identifier’s landing page always resolves to the most recent version, while earlier versions remain permanently accessible via the article’s Info/History tab.
This is a deliberate design choice, not a technical limitation. openRxiv has stated that “opinions differ on whether each version of a preprint should have its own DOI,” and different repositories take different approaches — but for bioRxiv and medRxiv, version-specific DOIs are not issued. A revision is only assigned a brand-new DOI if its content has changed so substantially that the author submits it as an entirely new manuscript rather than a revision.
Because Crossref registration is not instantaneous, a newly posted preprint’s DOI URL can take up to 24 hours to resolve. During that narrow window, linking directly to the bioRxiv or medRxiv article page — rather than the DOI — is the more reliable option for time-sensitive sharing.
| Attribute | Detail |
|---|---|
| Current DOI prefix (from 1 Dec 2025) | 10.64898 |
| Legacy DOI prefix (CSHL era, pre-Dec 2025) | 10.1101 |
| Suffix format (post 11 Dec 2019) | YYYY.MM.DD.###### — embeds author-approval date |
| DOIs per article | One; always resolves to the latest version |
| Version-specific citation | Append version number to the article URL, e.g. …/10.1101/2019.12.11.123456v2 |
| bioRxiv ISSN (electronic) | 2692-8205 (NLM Catalog / ISSN Portal) |
| medRxiv launch | June 2019, spun off from bioRxiv for clinical and health-science research |
| Registration agency | Crossref |
How do you cite bioRxiv v1 vs v3 correctly?
Because a single DOI serves every version, citing “v3” requires more than the bare DOI. bioRxiv’s own FAQ gives the format directly: cite the DOI, then append the version-specific URL if a particular version matters to the claim being made.
The standard citation format is: Author AN, Author BT. Year. Title. bioRxiv doi: 10.1101/2019.12.11.123456. To pin the citation to version 2 specifically, this becomes: doi: 10.1101/2019.12.11.123456 version 2, https://www.biorxiv.org/content/10.1101/2019.12.11.123456v2. The same logic applies to medRxiv and to articles carrying the newer 10.64898 prefix.
Version specificity matters most when:
- A reviewer or reader needs to see exactly what was public at the time a claim was made or a decision was taken
- An earlier version contained results, figures, or conclusions later revised or retracted
- A funder, journal, or regulator requires a dated, auditable snapshot of the manuscript (relevant to research administration compliance workflows)
Most citations across preprint servers are not version-specific — the majority of preprints only ever have a single version — so appending a version tag is the exception, applied only when precision genuinely matters. The National Institutes of Health (NIH) explicitly encourages citing preprints in grant applications, stating in its policy on Reporting Preprints and Other Interim Research Products that it “encourages investigators to use interim research products, such as preprints, to speed the dissemination and enhance the rigor of their work.”
How does the preprint DOI link to the published journal article?
When a preprinted manuscript is subsequently accepted by a journal, openRxiv’s matching algorithms detect the connection and update the preprint’s Crossref DOI metadata to point to the journal article’s DOI. This link typically appears within approximately two weeks of formal journal publication, and the corresponding author receives a confirmation request by email.
This bidirectional metadata relationship is what allows citation trackers, institutional repositories, and CRIS systems to treat the preprint and the published article as related outputs of the same research rather than duplicate records. Ideally the journal’s own DOI record reciprocally references the preprint, though bioRxiv’s FAQ notes this does not always happen in practice — a gap that research administrators should check for when auditing an author’s output list.
A related, often underappreciated fact: roughly two-thirds of bioRxiv preprints go on to be published in a peer-reviewed journal, according to meta-research published in eLife (Abdill & Blekhman, 2019) — meaning the DOI-linking mechanism is relevant to the majority of postings, not a rare edge case.
Common questions about bioRxiv and medRxiv DOIs
Does bioRxiv have a DOI?
Yes. Every bioRxiv and medRxiv preprint is assigned a DOI registered with Crossref once it clears screening. This makes preprints citable and part of the permanent scientific record, indexed by Google Scholar, Crossref, Semantic Scholar, Europe PMC, and the Preprint Citation Index in Web of Science.
How do you cite a bioRxiv preprint?
Cite the author list, posting year, title, and the DOI — for example, “bioRxiv doi: 10.1101/2019.12.11.123456.” To cite a specific version, append the version-specific URL after the DOI, following the format published on bioRxiv’s official FAQ page.
Is it okay to cite bioRxiv preprints?
Yes, with a caveat: bioRxiv states plainly that manuscripts “receive DOIs and thus are citable,” but they are not peer-reviewed at the time of posting. Readers and citers should note the preprint status explicitly and check whether a peer-reviewed version now exists before relying on it as a final source.
Do all preprint versions have their own DOI?
No. openRxiv assigns one DOI per article, shared by every revision. The DOI resolves to the newest version by default; earlier versions stay accessible through the article’s history tab and are cited by adding a version tag to the URL, not by obtaining a separate identifier.
Why this matters for research administrators
For institutions managing CRIS records, REF-style output audits, or funder compliance reporting, the DOI-prefix transition and the single-DOI-per-article model both have practical consequences. Output lists built before December 2025 will show 10.1101 DOIs; anything posted afterward will show 10.64898 — both are equally valid, permanent identifiers, and neither supersedes the other. Automated deduplication or metadata-harvesting scripts that pattern-match on the “10.1101” prefix should be updated to also recognise 10.64898, or they risk silently dropping newly posted preprints from institutional repositories.
The version-tracking model also has implications for research integrity workflows: because withdrawal notices and corrections are recorded against the same DOI rather than issued as a new identifier, institutions monitoring compliance should check an article’s Info/History tab — not just its DOI — before citing it in a report. As preprints continue to be formally recognised in NIH and UKRI reporting frameworks, treating the DOI as a static citation string, disconnected from its version history, is no longer sufficient practice for accurate scholarly record-keeping.
