medRxiv (pronounced “med-archive”) is a free, open-access preprint server for clinical and health-related research. Authors post manuscripts — before, during, or independent of journal peer review — where they receive a timestamp, a citable DOI, and public visibility within a few business days. Because clinical findings can influence patient care or public-health decisions if misread as settled, vetted guidance, medRxiv layers additional screening on top of the basic model used by its life-sciences sibling bioRxiv. This guide covers what medRxiv is, who runs it today, what it will and won’t accept, how its screening and licensing work, and how it relates to journal publication. For the underlying concept, see CASRAI’s definition of preprint; for the fuller landscape of servers, see Preprint Servers Explained and How to Choose a Preprint Server. If you are ready to submit, see the step-by-step bioRxiv/medRxiv submission guide. For a side-by-side comparison with bioRxiv, see bioRxiv vs. medRxiv.
Origins and who runs medRxiv today
medRxiv launched in June 2019, several years after bioRxiv, following a restricted pilot in which bioRxiv accepted a limited set of clinical preprints. It was created as a managed partnership between Cold Spring Harbor Laboratory (CSHL), Yale University School of Medicine, and the medical publisher BMJ — a deliberately broader governance group than bioRxiv’s, reflecting the greater sensitivity of clinical content.
medRxiv’s operations, along with bioRxiv’s, were spun out of CSHL into openRxiv, an independent non-profit organization, as of March 11, 2025. openRxiv now operates both servers day to day. This is a genuinely current detail worth getting right: medRxiv is not run on the Center for Open Science’s OSF Preprints infrastructure, which is a separate platform hosting a number of other, unrelated subject-specific preprint servers. medRxiv has always run on its own dedicated platform, first built and maintained under the CSHL/Yale/BMJ partnership and now under openRxiv.
Scope: what kind of work goes on medRxiv
medRxiv is scoped narrowly to clinical and health research: randomized and non-randomized clinical trials, epidemiology, cohort and observational studies, public-health research, and related human-health work. This is a meaningfully tighter scope than bioRxiv’s broad life-sciences remit — basic and preclinical biology, even where the eventual application is medical, generally belongs on bioRxiv instead, and medRxiv’s screening team will redirect a manuscript submitted to the wrong server rather than simply reject it.
Several content types are explicitly excluded from medRxiv, not merely discouraged: case reports and case series, preclinical or veterinary studies, narrative reviews, editorials and opinion pieces, hypotheses presented without supporting data, and standalone guidelines, policy statements, or protocols with no accompanying results. Case reports are excluded specifically because a single, unreplicated clinical observation carries an outsized risk of being misread as clinical guidance — a risk medRxiv’s screening standard weighs explicitly in a way bioRxiv’s broader scope does not need to.
Screening: stricter than bioRxiv’s baseline model
Like every reputable preprint server, medRxiv screening is not peer review, and openRxiv does not describe it that way. Every submission goes through a basic check for plagiarism, offensive or non-scientific content, and biosecurity/dual-use risk — the same baseline bioRxiv applies. medRxiv goes further, because of the greater potential for public or patient harm if a clinical preprint’s findings are acted on before they’ve been vetted:
- Ethics/IRB documentation: submissions involving human subjects must include a statement of oversight and approval by an Institutional Review Board, Ethics Committee, or equivalent body.
- Clinical trial registration: clinical trials must carry a trial ID from an ICMJE-recognized trial registry (such as ClinicalTrials.gov).
- Causal-language caution: authors are cautioned against asserting causal claims without a robust causal-inference framework to support them; a manuscript that overstates causality relative to its study design may be sent back or declined.
As with bioRxiv, every medRxiv preprint carries a visible disclaimer stating it has not been certified by peer review and should not be used to guide clinical practice or reported in the press as established information. Clinicians, journalists, and the public are expected to treat medRxiv content as unreviewed, preliminary research.
Licensing choices
Authors choose a license at posting time from the same set bioRxiv offers: CC BY, CC BY-NC, CC BY-ND, CC BY-NC-ND, or CC BY-NC-SA, plus a no-license “all rights reserved” option for authors not ready to release their work more openly. As with bioRxiv, a January 2026 openRxiv policy change lets authors loosen — but not tighten — a posted preprint’s license after the fact, without filing a new version, provided the preprint is less than two years old.
DOIs and versioning
Each medRxiv article receives a single DOI that always resolves to its most recent version; older versions remain permanently accessible at version-specific URLs, and authors can post a revised version at any point rather than editing the original in place. medRxiv shares bioRxiv’s DOI-prefix history: preprints posted before December 1, 2025 carry a 10.1101 prefix, while preprints posted from that date onward, coinciding with the CSHL-to-openRxiv operational transition, use a new 10.64898 prefix. Existing 10.1101 DOIs continue to resolve normally and were not reissued. To cite a specific version, openRxiv’s guidance is to append the version number to the DOI in citation text or link the version-specific URL directly. See Preprint Version Numbers: Citation and Tracking Across Servers for the full mechanics.
How a medRxiv preprint relates to eventual journal publication
Posting to medRxiv does not compete with journal publication. Most journals, including those following ICMJE recommendations, do not treat a preprint as prior or duplicate publication provided the author discloses it at submission; see Preprint Posting and the “Prior Publication” Conflict for the conditions that generally apply. A medRxiv preprint’s page is typically updated to link forward to the version of record once it is published in a journal, via Crossref’s preprint-to-published relationship metadata. Author lists occasionally change between the preprint and the published version; see Do Author Lists Change Between a Preprint and Its Published Version? for how often that happens and why.
What medRxiv does not do
medRxiv has no editorial board issuing accept/reject decisions on scientific merit beyond its screening criteria, no formal retraction mechanism equivalent to a journal’s, and no guarantee that a posted manuscript will ever be submitted to or accepted by a journal at all. Because clinical preprints carry higher public-harm potential than most basic-science preprints, medRxiv’s disclaimer and screening exist precisely to manage that gap between “posted” and “peer reviewed” — but they don’t eliminate it. openRxiv posts withdrawal notices rather than silently removing content, and version history stays visible when a correction or revision is needed. See How Preprint Servers Handle Misconduct and Retraction and Disadvantages of Preprints for Authors for a fuller look at these tradeoffs, and CASRAI’s guide to citing a preprint for how to format a citation that makes clear the work is unreviewed.
How medRxiv compares to other preprint servers
medRxiv is the standard choice for clinical and health research specifically; for basic or preclinical life-sciences work, bioRxiv is the equivalent, broader-scope server — see CASRAI’s full bioRxiv vs. medRxiv comparison for the complete scope and screening breakdown between the two. For a discipline-specific server built on the same screening-not-peer-review model outside biomedicine, see What Is EarthArXiv?, the earth-science equivalent, governed by its own community Advisory Council and hosted via the University of California’s California Digital Library rather than openRxiv.
Frequently asked questions
Is medRxiv peer reviewed?
No. medRxiv preprints go through screening for plagiarism, offensive or non-scientific content, biosecurity risk, and (unlike bioRxiv) additional clinical-specific checks such as IRB documentation and trial registration — not peer review. Every posting carries a disclaimer stating it has not been certified by peer review and should not guide clinical practice.
Who owns and runs medRxiv?
medRxiv launched in June 2019 as a partnership between Cold Spring Harbor Laboratory, Yale University School of Medicine, and BMJ. Since March 11, 2025 it has been operated by openRxiv, an independent non-profit spun out of CSHL’s bioRxiv/medRxiv operations. It is not run on OSF, the Center for Open Science’s separate preprint infrastructure.
What is the difference between medRxiv and bioRxiv?
medRxiv is scoped specifically to clinical and health-related research and applies extra screening — IRB documentation, ICMJE-recognized trial registration for clinical trials, and caution around causal-language claims — that bioRxiv’s broader life-sciences scope does not require. See the full bioRxiv vs. medRxiv comparison.
Does medRxiv accept case reports?
No. Case reports and case series are explicitly excluded, along with preclinical or veterinary studies, narrative reviews, editorials/opinion pieces, hypotheses without supporting data, and standalone guidelines or protocols without results.
Can I still publish in a journal after posting to medRxiv?
In almost all cases, yes. Most journals and publishers, including those following ICMJE recommendations, do not treat a preprint as prior publication, provided you disclose it at submission. See Preprint Posting and the “Prior Publication” Conflict for specifics and exceptions.
Does a medRxiv preprint have a DOI I can cite?
Yes. Every medRxiv preprint receives a DOI at posting, which resolves to the latest version; older versions remain accessible via version-specific URLs. See Preprint Version Numbers: Citation and Tracking Across Servers for how to cite a specific version.
Is medRxiv free to use?
Yes. There is no fee to post a preprint or to read one, and no institutional subscription is required to access content.
