bioRxiv: the preprint server for biology

Richard Sever; Samantha J. Hindle; Ted Roeder; Sol Fereres; Olaya Fernández‐Gayol; Sanchari Ghosh; Martina Proietti Onori; Emma Croushore; Kevin J. Black; Linda K. Sussman; Janet Argentine; Wayne Manos; Marisol Muñoz; Josh Sinanan; Tracy K. Teal; John R. Inglis

doi:10.1101/833400

bioRxivFeb 26, 2026Scientific Communication & EducationPreprint

bioRxiv: the preprint server for biology

Richard Sever iD, Samantha J. Hindle iD, Ted Roeder iD, Sol Fereres, Olaya Fernández‐Gayol iD, Sanchari Ghosh, Martina Proietti Onori, Emma Croushore, Kevin J. Black iD, Linda K. Sussman, Janet Argentine, Wayne Manos, Marisol Muñoz, Josh Sinanan, Tracy K. Teal, John R. Inglis iD

PDF DOI Data

In-depth full-text digest available · Jump to digest ↓

Abstract

The traditional publication process delays dissemination of new research, often by months, sometimes by years. Preprint servers decouple dissemination of research papers from their evaluation and certification by journals, allowing researchers to share work immediately, receive feedback from a much larger audience, and provide evidence of productivity long before formal publication. Launched in 2013 as a non-profit community service, the bioRxiv server brought preprinting to the life sciences and recently posted its 310,000th manuscript. The server now receives around ten million views per month and hosts papers spanning all areas of biology. Initially dominated by evolutionary biology, genetics/genomics and computational biology, bioRxiv has been increasingly populated by papers in neuroscience, cell and developmental biology, and many other fields. bioRxiv and its sister server, medRxiv, also played a critical role during the pandemic, rapidly disseminating new discoveries in immunology, virology and epidemiology related to the SARS-CoV-2 virus and its effects. Changes in journal and funder policies that encourage preprint posting helped drive adoption, as did the development of bioRxiv technologies that allow authors to transfer papers easily between bioRxiv and journals. A recent user survey found that 30% of authors post their preprints weeks to months before submitting to a journal whereas 55% post around the time of journal submission. Authors are motivated by a desire to share work early; they value the feedback they receive and very rarely experience any negative consequences of preprint posting. Rapid dissemination via bioRxiv is also encouraging new initiatives that experiment with the peer review process and the development of novel approaches to literature filtering and assessment.

research focus: database / resource

In-depth summary

Claude Opus 5.5 · max effort · 8 Nov 2026 · from the full text of version 2 (26 Feb 2026) · CC BY 4.0

Since its 2013 launch, bioRxiv has posted more than 310,000 life-science preprints; about 80% later reach journals, and surveyed authors rarely report negative consequences of posting.

Coupling dissemination to journal evaluation delays the sharing of research by months or years, and earlier attempts to bring arXiv-style preprinting to biology failed amid publisher opposition and scant interest from biologists. The article documents how bioRxiv screens, hosts, links and archives preprints, backed by platform usage statistics and a 2023 survey of more than 7000 bioRxiv and medRxiv users.

Key results

ShownbioRxiv expanded from a handful of monthly postings after its 2013 launch into a high-volume, international server whose content broadened from genomics, bioinformatics and evolutionary biology toward neuroscience and other fields.
More than 310,000 first submissions, with over 4000 per month in 2025; >7 million abstract views and ~4 million PDF downloads per month; authors from 209 countries (most often USA, UK, Germany, China); 26% of preprints revised (8% twice, 4% three or more times). · Abstract; bioRxiv by the numbers
ShownMost bioRxiv preprints are later published across a wide range of journals, and withdrawals are rare, occurring at a rate similar to retraction of journal articles.
Title/author matching finds ~80% published after a 3-year window (possibly an underestimate; consistent with arXiv), in more than 5000 journals; mean posting-to-publication interval 256 days (median 204 days); 500 withdrawals during 2018 to 2025, a 0.17% rate vs ~0.2% of journal articles retracted. · bioRxiv by the numbers
SuggestedAmong survey respondents, most post preprints around or before journal submission, mainly to raise visibility and claim priority, and the vast majority report no adverse consequences.
Survey of >7000 bioRxiv and medRxiv users (2023): 55% post in the week of first-choice journal submission, 30% weeks to months earlier (17.6% 1-4 weeks; 12.5% >1 month); motives include awareness (82%) and priority (56%); 89% saw no negative effects, 2% believed a competitor's head start cost them a specific journal, 4% felt journal choice narrowed. · Abstract; The 2023 bioRxiv survey
SuggestedPublic on-site discussion of preprints is sparse; respondents receive feedback mostly through private email, conversations and social media, so a stated appetite for online comments is only partly met.
Platform data: 4% of papers display linked peer reviews, 4% carry on-site comments and just over 2% have third-party community reviews; survey: feedback came via email (36%), conversation (34%), Twitter/X (32%) or bioRxiv comments (9%); 46% strongly want online comments, yet 58% have never given preprint feedback. · bioRxiv by the numbers; The 2023 bioRxiv survey
ShownbioRxiv and medRxiv became major channels for rapid release of SARS-CoV-2 and COVID research, with submissions peaking in May 2020 and later declining to a low plateau.
Dedicated COVID collection holds >30,000 papers; the first SARS-CoV-2 preprint appeared on bioRxiv on January 19th, 2020; combined bioRxiv plus medRxiv COVID submissions exceeded 2000 in the peak month, May 2020, then fell through 2024 to ~100/month in 2025. · bioRxiv in the pandemic

Methods, data and reproducibility

Descriptive account of bioRxiv's platform (BenchPress submission, Highwire JCore display, two-stage staff and scientist-Affiliate screening, PubMed/Crossref journal matching, bioRxiv-to-journal (B2J) and journal-to-bioRxiv (J2B) transfers) plus server statistics from api.biorxiv.org and a 2023 survey of more than 7000 bioRxiv and medRxiv users.

bioRxiv posting records, 2013 to 2025: monthly new and revised submissions, revision counts, subject categories, licences, author countries and institutions · n = more than 310,000 first submissions
Withdrawn bioRxiv preprints (author- or server-initiated), 2018 to 2025 · n = 500 papers
Preprint-to-journal publication links found by scripts matching titles and authors against PubMed and Crossref, with destination journals and publication intervals
Monthly site usage: abstract views, PDF downloads and full-text HTML views
Per-preprint engagement: linked journal or review-service peer reviews, on-site Disqus comments and third-party community reviews
COVID preprint collection tagged across bioRxiv and medRxiv, with monthly submission counts · n = >30,000 papers
2023 survey of bioRxiv and medRxiv users on authoring tools, posting motives and timing, feedback channels and career effects (institutional review board exempt) · n = >7000 bioRxiv and medRxiv users

Methods & tools: BenchPress submission system (Highwire Press), customized for preprint screening and transfer, Highwire JCore display platform, Similarity Check plagiarism detection, Hypothesis annotation, used to display linked peer reviews and automated outputs, Disqus commenting plug-in, Crossref DOI registration and API, PubMed and Crossref title/author matching scripts, Solr site search, Portico long-term archiving, Sciscore (research resource identifier extraction), Sciencecast (AI-generated summaries), B2J/J2B ZIP-over-FTP manuscript transfer adapted to Manuscript Exchange Common Approach (MECA)
Data & accessions: http://api.biorxiv.org/
Reproducibility gaps: De-identified survey data are provided via a cited reference, with no URL in the text; API metrics update in real time, so reported snapshot values will drift; preprint-journal matching scripts are not released.

Limitations (6)

Survey respondents were self-selected with uneven geographic representation, limiting generalization; a 9% rate of feedback via bioRxiv comments, against 4% of papers carrying comments, implies sampling bias, and 2023 data likely understate current large language model use (10%). · stated by the authors
The ~80% journal publication rate may be an underestimate, since matching misses preprints whose titles or author lists changed before publication and those still under journal review. · stated by the authors
Screening by staff and scientist Affiliates is a quick, coarse filter rather than peer review; it neither guarantees nor certifies content, so readers must judge scientific validity themselves. · stated by the authors
Ensuring withdrawal status propagates to third-party indexers is still under investigation, and scientific screening plus parts of submission metadata entry remain manual, with AI assistance only anticipated. · stated by the authors
Comparing the 0.17% bioRxiv withdrawal rate (2018 to 2025) with ~0.2% of journal articles retracted equates mostly author-initiated withdrawals with formal retractions over possibly unmatched time windows, so the similarity is only indicative. · noted in this summary
Survey response rate and per-question denominators are not reported and percentages lack uncertainty estimates; with most authors holding openRxiv or Cold Spring Harbor Laboratory roles, metrics and survey are operator-produced without independent audit. · noted in this summary

Read this if you need a citable reference on bioRxiv's screening, indexing and journal-transfer infrastructure, its publication and withdrawal rates, or survey data on why and when life scientists post preprints.