Since 2023, eLife has run one of the most closely watched experiments in scholarly publishing: it no longer makes a traditional accept/reject decision on a manuscript before publishing it. Instead, papers selected for peer review are published as a Reviewed Preprint — the manuscript, a set of public peer reviews, and an editorial eLife assessment — without eLife itself deciding whether the paper is ‘accepted.’ That decision is left to readers, funders, and the author’s own institution. This guide explains how the model actually works end to end, how it differs from both traditional peer review and an unreviewed preprint, and what its reception has looked like since launch.
What a Reviewed Preprint actually is
A Reviewed Preprint is the version of a manuscript that eLife publishes on its own site once peer review is complete, bundled with three things: the preprint itself, one or more signed Public Reviews written by the peer reviewers, and a short editorial eLife assessment that characterizes the work using a controlled vocabulary. Authors may also publish a response to the reviews. Every Reviewed Preprint receives its own DOI and is citable in that form — it does not need to wait for a later ‘final’ version to be referenced.
This is eLife’s implementation of what it originally called the Publish-Review-Curate (PRC) model: publish the preprint, review it openly, then let the author and the community curate what happens next, rather than a journal gatekeeping publication itself.
How the process works, step by step
- Submission. Authors submit directly to eLife, or transfer a preprint already posted on bioRxiv or medRxiv.
- Editorial screening. eLife’s editors — practicing researchers, not full-time journal staff — assess each submission against the journal’s scope and quality bar. Papers that don’t clear this bar are desk-rejected without going to review; per eLife’s own 2025 figures, the average time to a desk-rejection decision was about 8 days.
- Selection for review. Papers that pass screening are sent out for peer review. In eLife’s most recent reporting (2025), roughly a third of submissions were sent for review — a higher proportion than in prior years, which eLife attributes to a smaller but more self-selected submission pool since a 2024 policy change described below.
- Peer review. Reviewers write a Public Review of the manuscript rather than a private, confidential recommendation to an editor. A publication fee applies once a paper enters this stage (eLife lists it at $3,750 as of this writing, with waivers available); this replaces the traditional ‘free to submit, pay only on acceptance’ fee structure precisely because there is no acceptance gate to pay at.
- The eLife assessment. The handling editor, working with the reviewers, writes a short assessment summarizing the significance of the findings and the strength of the supporting evidence, using standardized descriptive terms (significance ranging from ‘landmark’ to ‘useful’; strength of evidence ranging from ‘exceptional’ to ‘inadequate’). The intent is to give a reader a fast, comparable signal without collapsing the whole review into a single accept/reject verdict.
- Publication as a Reviewed Preprint. The manuscript, Public Reviews, and eLife assessment are published together. This is the point at which, functionally, the paper becomes eLife’s citable output for that manuscript — there is no separate ‘acceptance’ event before this happens.
- Revision, at the author’s discretion. Authors decide whether and how to revise in response to the reviews. A revised version can be published as an updated Reviewed Preprint with an updated assessment. Authors may also choose to publish a Version of Record once they consider the review process complete, but doing so is optional, not a required final gate.
How this differs from traditional peer review
In a conventional journal workflow, peer review happens before anything is public, reviewers report privately to an editor, and the editor makes a binary decision (accept, reject, or revise-and-resubmit) that determines whether the work is published under that journal’s name at all. Readers only ever see the version the editor ultimately approved.
eLife’s model inverts the visibility and moves the decision point: review happens in public, on a version the author already chose to make available, and the editor’s output is an assessment rather than a verdict. The author — not the journal — decides what happens after the reviews are public, including whether to treat the current Reviewed Preprint as the final word.
See Peer Review: What It Is, How It Works, and Why It Matters for the general mechanics eLife’s model departs from, and Single-blind vs double-blind vs open peer review for how eLife’s fully open, signed reviews compare with other openness models.
How this differs from an unreviewed preprint
A plain preprint posted to a server like bioRxiv or medRxiv has had no independent expert review at all — it is the author’s own manuscript, unvetted. A Reviewed Preprint has been through eLife’s editorial screening and expert peer review, and carries reviewer commentary and an editorial assessment alongside it. The practical difference for a reader is significant: a Reviewed Preprint tells you what qualified reviewers found strong or weak about the work, in public, attached to the object itself, rather than leaving that judgment entirely to the reader or to a later publication decision elsewhere.
For background on the format itself, see How to Choose a Preprint Server and Preprints vs. Peer-Reviewed Articles: What’s the Real Difference?; for the mechanics of another journal running a comparably open model, see F1000Research and Open Peer Review: How the Model Works.
Current status and reception
eLife’s own three-year update on the model (published on elifesciences.org) reports a median time from submission to Reviewed Preprint publication of around 96 days, compared with an average delay of roughly 250 days between preprint posting and eventual journal publication elsewhere — the speed gain the model was designed to produce.
The most significant disruption to the model’s reception came in late 2024: eLife did not comply with Clarivate’s Web of Science indexing requirements — largely because the journal’s output no longer fits a framework built around a single accept/reject-defined article — and was delisted, losing its Journal Impact Factor. eLife’s own reporting attributes a substantial drop in submissions to this change: from 6,396 submissions in 2024 to 2,853 in 2025, a fall of roughly 55%, with the size of the decline varying sharply by country and field. At the same time, the proportion of submissions sent out for full peer review rose (from about 27% to about 35%), which eLife frames as evidence that the remaining submission pool is more self-selected toward papers suited to the model rather than a sign of declining interest per se.
Reception has been uneven by institutional and disciplinary context. At research-intensive institutions and funders already invested in open science, the model is well regarded and increasingly built into assessment practice. In fields and institutions still anchored to impact-factor-based evaluation — and in disciplines with little preprint culture, such as much of clinical medicine — a ‘published as a Reviewed Preprint’ outcome is harder to explain to a tenure or promotion committee than a conventional ‘accepted at Journal X.’
Funder support has grown alongside this friction. eLife reports that funders including the Howard Hughes Medical Institute (HHMI), Wellcome, the Bill & Melinda Gates Foundation, and the Knut and Alice Wallenberg Foundation have stated they will accept reviewed preprints in research evaluation processes such as grant applications and progress reports. Wellcome announced a further £2.4 million investment in eLife’s infrastructure (eLife Pathways) in March 2026. This funder backing connects directly to the broader research-assessment reform conversation represented by DORA and CoARA, both of which call for evaluating research on its own merits rather than by the venue or impact factor it was published under — see DORA and CoARA in Practice and DORA vs CoARA for how that reform agenda is being operationalized at the institutional level.
Practical implications for authors and research offices
- There is still a gate — it just moves. eLife still declines papers; it simply does so through desk rejection before review rather than through a post-review ‘reject’ decision. A paper that reaches Reviewed Preprint status has cleared eLife’s editorial screening and been reviewed by named experts, even though no single up-or-down verdict follows.
- Check funder and institutional policy before relying on a Reviewed Preprint for reporting. Some funders explicitly count reviewed preprints as peer-reviewed outputs for grant reporting and assessment; others’ policies have not caught up to the model. Research offices advising authors on where to submit, or compiling outputs for a progress report, should confirm current guidance with the specific funder rather than assuming either way.
- Don’t assume Journal Impact Factor will be available. Since the 2024 Web of Science delisting, eLife output has no Journal Impact Factor. For authors or committees in fields where JIF is still a formal or informal evaluation criterion, this is a real, current tradeoff to weigh against the model’s speed and transparency benefits.
- A Reviewed Preprint is a legitimate citable output on its own. It does not require a subsequent Version of Record to be referenced in a CV, grant application, or another paper’s reference list — it carries its own DOI from the point of publication.
Frequently asked questions
Does eLife still reject papers before review?
Yes. eLife’s editors screen every submission and desk-reject papers that don’t meet the journal’s scope and quality bar before they ever go to peer review — per eLife’s 2025 figures, that decision takes about 8 days on average. What eLife has removed is the second gate: a post-review accept/reject decision. Once a paper is selected for review and reviewed, it is published as a Reviewed Preprint regardless of how favorable the reviews are.
Can authors revise after a Reviewed Preprint is published?
Yes, and the choice is the author’s. Authors can submit a revised version responding to the Public Reviews, which is published as an updated Reviewed Preprint with an updated eLife assessment. Authors can also choose to publish a Version of Record once they consider the process finished, but nothing requires them to do either.
Is a Reviewed Preprint peer reviewed for the purposes of a funder mandate?
It has been through expert peer review — the Public Reviews are written by reviewers eLife’s editors selected for their expertise. Whether a specific funder’s open-access or reporting policy formally counts a Reviewed Preprint as satisfying a ‘peer-reviewed’ requirement depends on that funder’s own current policy language; several major funders (HHMI, Wellcome, Gates, Wallenberg) have stated they accept reviewed preprints in evaluation, but this is not yet universal, so it is worth confirming directly against the relevant funder’s guidance rather than assuming.
What is an eLife assessment?
A short, structured editorial summary published alongside every Reviewed Preprint, written by the handling editor with input from reviewers. It characterizes the significance of the findings (on a scale eLife describes as running from ‘landmark’ to ‘useful’) and the strength of the supporting evidence (‘exceptional’ to ‘inadequate’), giving readers a fast, comparable signal without reducing the review to a single accept/reject outcome.
What happened to eLife’s impact factor?
eLife did not comply with Clarivate’s Web of Science indexing requirements in late 2024 and was delisted, meaning it no longer receives a Journal Impact Factor. eLife has framed this as a consequence of the journal’s output no longer fitting a model built around discrete, journal-decided ‘accepted articles,’ and reports it as a contributing factor in a subsequent drop in submission volume.
Sources: elifesciences.org (About/Peer Review pages and eLife’s own three-year model update, ‘The eLife Model: Three-year update,’ Inside eLife), consulted July 2026. Figures on submission volume, review rates, desk-rejection time, and funder commitments are eLife’s own self-reported figures; treat them as eLife’s account of its own model rather than independently audited statistics.
