Desk Rejection Rate by Journal: Published Figures With Sources (2026)

What counts. A row needs a desk-rejection rate that the journal, its publisher or society, or a peer-reviewed study of that journal published, and we read the figure on the source itself. The rate is the share of submissions that editors rejected before external peer review. We accepted three definitions and record which one each source uses:

• Desk rejected: the source uses that term.
• Rejected without review: the source says rejected without external review, before peer review, at initial screening, or summarily.
• Not sent for review: the source gives only the share sent out for review, so the rate is 100% minus that share. This can include a few withdrawn or still-pending papers.

How figures were derived. When a source gives counts instead of a percentage (for example, 818 summary rejections out of 2,114 decisions at the American Economic Review), we divide them and mark the row "computed from published counts". The CSV and JSON carry the arithmetic and the denominator for every row.

What we left out. Estimates, journal-finder and ranking sites, SciRev, BioxBio, writing-service blogs and secondary summaries, even when the number looked plausible. Figures reported only for one article type or a subset of submissions are kept only when the subset is named, and the row says so. Ten figures that the research pass had read on their sources were still dropped at review, for example where a source's own counts did not add up, a denominator was ambiguous, the figure covered a whole publisher rather than one journal, or the data were from 2010.

When we checked. All 88 rows were read on 2 October 2026. Each row shows that date, and the change log records later edits. Send corrections to erik@manusights.com.

• Self-selected sample. Journals that publish editor's reports (economics, sociology, political science and society medical journals) are over-represented, and the most selective general journals mostly publish nothing. The 50% median describes this table, not all journals.
• Definitions and denominators differ. Some rows divide by new submissions, some by decisions made in the year, and a few by one article type. JCI's own page notes that its numerator and denominator come from different cohorts. Compare rows within a field before comparing across fields.
• Rates move. The American Political Science Review reported desk rejection of 18.8% in 2007 to 2012 and 72.0% under its editorial team from 2024. Rows older than six years are flagged as older figures; Nature and Nature Communications are from 2015 to 2017 because no newer published figure exists.
• Bounds are bounds. "More than half", "fewer than 10%" and "up to 50%" are printed as the source wrote them and sorted at the stated bound.

Manusights (2026). Desk-rejection rates by journal (Version 2026-10) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.23353184

The Zenodo record holds this edition's files, fixed and versioned. For the 18 journals whose publishers' terms forbid compiling their figures, the JSON file lists the journal and publisher but not the figure; the figure is on this page with its source. Each journal row has its own link: add the journal's anchor to the address, for example https://manusights.com/blog/desk-rejection-rate-by-journal#nature for Nature. When you quote a single journal's figure, cite that journal's source as well; the link is in its row.

Several numbers that circulate as "the" desk-rejection rate for well-known journals do not match what the journals or studies of them report:

• PLOS ONE reported 30.8% of submissions rejected without peer review for January to June 2023, up from between 21% and 23% in each of the six half-years before it. The 15% to 20% figure still quoted for PLOS ONE is out of date.
• Science Advances sent 25.9% of regular submissions past the editorial stage in 2015 to 2020, which puts its pre-review filter (74.1%) close to Science's (82.6%), not at the 40% often quoted.
• eLife sent 1,013 of its 2,853 submissions in 2025 for review (64.5% not reviewed), lower than the 80% to 85% screen it is often credited with.
• Nature rejected 79.1% of submissions outright at the first editorial decision in 2015 to 2017, higher than the 60% that many tables still show.
• Physical Review B reported rejecting about a quarter of submissions without external review in 2018, not 10% to 15%.

The same pattern holds in reverse for journals that are assumed to be easy: Cyberpsychology reported an 81% desk-reject rate in 2024 and Croatian Medical Journal about 85%. Small and regional journals can screen as hard as flagships.

These are the estimates this page showed before the verified table existed. They are Manusights planning estimates, not journal data, and none of these journals publishes a current desk-rejection rate we could find. They are kept because readers and other sites have cited them; use them as a rough tier signal only, and prefer the verified table wherever a journal appears there.

What changed from the earlier version of this page. Ten earlier estimates were replaced by published figures: Nature, Science, JAMA, The BMJ, Nature Communications, PLOS ONE, eLife, PNAS, Science Advances and Physical Review B. Four rows were removed: Angewandte Chemie ("moderate") and Advanced Materials ("high") had no figure, Nature Chemical Biology had none to give, and the PLOS Medicine estimate conflicted with the journal's own statement about how few initial submissions it invites. The "overall acceptance" and "first decision" columns were removed because they were unsourced; see journal acceptance rates and journal review timelines for sourced, labelled figures.

In our analysis of desk-rejection-risk reviews, the specific failure pattern is not "the journal is too selective." It is that the manuscript gives the editor a quick reason to stop: the abstract promises more than the figures show, the scope reads like a different journal, or the reporting package is incomplete before review starts.

We see three repeat patterns in Manusights review data:

• Scientific Reports scope confusion: authors treat Scientific Reports as "easier Nature" instead of a soundness-led multidisciplinary journal. The safer test is whether the methods, ethics, data availability, and reporting package are complete enough for review.
• Selective-journal pitch mismatch: authors target Nature, Science, NEJM, JAMA, Cell, or Lancet with a cover letter that summarizes the study but does not explain why the result changes the journal's field-wide conversation.
• Fallback-plan delay: authors wait for the desk rejection before choosing the next journal. For high-filter targets, the fallback should be named before upload so the abstract and cover letter can be retargeted quickly.

Having reviewed thousands of manuscripts through our manuscript readiness check, many targeting the journals listed on this page, we can add context that the numbers alone don't provide.

The single most consistent predictor of desk rejection across all journals: the abstract and cover letter are calibrated for a different journal's audience. A paper written for a specialty journal and submitted to a broad-science journal (Nature Communications, PNAS, Science Advances) without reframing the significance for a general audience gets desk-rejected.

A paper written for a general audience and submitted to a specialist journal (JCI, Water Research, ACS Nano) without deepening the technical detail gets desk-rejected. The science is the same in both cases. The framing determines the outcome.

The other pattern: authors who treat high desk-rejection rates as evidence that submitting is futile. At Nature Communications, which rejected 72.5% of submissions at the first editorial decision in 2015 to 2017, the papers that reach reviewers are not fundamentally different science. They are better-framed science. The cover letter, abstract, and first figure do most of the editorial work. Investing an hour rewriting those three elements for each target journal is the highest-return pre-submission investment we can recommend.

For journals with cascading systems (Nature/Nature Communications, Cell/Cell Reports, Science/Science Advances), desk rejection at the flagship is often a routing signal, not a quality judgment. A paper desk-rejected from Science with feedback about "insufficient breadth" is being told it belongs at Science Advances. Taking that feedback literally and adjusting the framing saves months.

A 72% desk-rejection rate at the American Political Science Review does not mean 72% of submitted papers are bad. It means the journal receives far more competent papers than it can publish and filters first for fit and contribution. A desk rejection from a highly selective journal is a selectivity judgment, not a quality verdict, and the same paper may be accepted at a field journal.

The most common desk rejection reasons across journals are:

1. Scope mismatch (the paper does not match what the journal publishes)
2. Significance insufficient for the journal tier
3. Methodology concerns visible from the abstract
4. Overclaimed conclusions relative to study design
5. Reporting incomplete (missing checklists, ethics, data availability)

Items 1 and 2 are about journal targeting. Items 3 through 5 are about manuscript quality. Both are checkable before submission.

At most selective journals, the desk decision is made on the first read of the abstract and figures, and the fast turnaround that journals report bears that out: the American Journal of Political Science reported a mean of 4.8 days to a desk rejection in 2025. Editors do not read the full methods section before deciding on triage. The abstract must communicate significance immediately, and the first figure must show the key result. If the significance is not obvious on that first read, the paper is likely to be returned regardless of the quality of the underlying work.

The manuscript readiness check evaluates your manuscript against your target journal's specific editorial standards in about two to three minutes. You get a readiness score, desk-reject risk signal, and the top issues with direct quotes from your paper. This catches the issues in categories 3 through 5 above (methodology, overclaiming, incomplete reporting).

For a full assessment, the manuscript readiness check provides citation-integrity checks against scholarly metadata sources, figure-level feedback, and a prioritized revision checklist. Citation risk is checked against CrossRef and PubMed. This catches problems that a manual check would take hours to identify.