Of 902 manuscripts submitted to one clinical journal in a single year, 552 never reached a reviewer. An editor returned them at screening, most within days, for reasons that had little to do with whether the science was correct. Those figures come from a Pakistan Journal of Medical Sciences editorial that collects screening data across several journals, and they are unremarkable. A 61% pre-review rejection rate is ordinary.
Authors read that outcome as a judgment on their work. It usually is not. A desk rejection answers a narrower question than peer review does: is this manuscript in this journal’s territory, and is it in a state where asking two overcommitted referees to spend six hours on it is defensible. A paper can be correct, careful, and months of honest work, and still fail that question in four minutes.
The screen is fast, repetitive, and mostly documented in public. Editors publish their criteria in author guidelines and editorial policies, then apply them in a fixed order. That makes the screen something you can run yourself, on your own manuscript, before you submit.
What is a desk rejection?
A desk rejection is an editorial decision to decline a manuscript without sending it out for external peer review. You will also see it called rejection without review, editorial rejection, pre-review rejection, or triage. The decision typically comes from a handling editor or the editor-in-chief, sometimes with input from an editorial board member, and it arrives days rather than months after submission.
The mechanism behind it is referee supply. Reviewing is unpaid and slow, and every manuscript sent out consumes a favor an editor will need again next month. Editors treat that supply as the resource they are managing, which is why they decline aggressively and quickly. Nature states the policy plainly: a substantial proportion of manuscripts are declined without being sent to referees so that authors can send them elsewhere without further delay.
That framing matters for how you read your own rejection letter. The editor was not asked whether your regression was specified correctly. They were asked whether this paper is worth a referee’s Saturday.
How common is desk rejection?
Common enough that it is the single most likely outcome of a submission at most selective journals, and the rate varies by roughly a factor of two between journal types.
Published figures give the shape of it. Parasites & Vectors reported a 39% pre-review rejection rate; JAMA Internal Medicine reported 78% in 2017. Both numbers appear in a 2022 editorial, “Top 10 reasons your manuscript may be rejected without review”, by Filipe Dantas-Torres. The clinical journal cited above screened out 61%. Broad, high-selectivity journals sit at the top of that range because the volume of plausible-but-not-for-us submissions is enormous; specialist journals that assess soundness rather than impact sit lower, because fewer submissions are off-territory to begin with.
Two variables drive most of the spread. The first is how much of the journal’s selectivity is enforced at screening versus at review. The second is scope breadth: a journal that publishes anything methodologically sound in a defined field rejects less at the desk than one selecting for general interest, because “not interesting enough for our readership” is a screening judgment while “not sound” usually is not.
The practical read: if you are submitting to a general or flagship journal, assume the screen is the hardest gate you face, and prepare for it first.
Why do editors reject papers without review?
Almost all desk rejections fall into three groups, and only one of them concerns the quality of your science.
The Dantas-Torres editorial lists ten recurring reasons: out of scope, local interest only, no novelty, inadequate study design, incomplete methodological detail, data availability problems, ethical issues, poor presentation, poor English, and plagiarism. Sort those into buckets and the pattern is clearer.
Fit. Out of scope, local interest, insufficient novelty for this readership. These are decisions about the journal, not about your data. A regional cohort study rejected as “of local interest” by a global journal may be exactly right for a national one.
Compliance and integrity. Missing ethics approval or consent statements, no trial registration, absent data availability statement, high text similarity, undisclosed prior publication. These are close to automatic. Many are checked before a human forms an opinion at all, and an editor who finds one rarely spends further time on the paper.
Presentation. Formatting that ignores the guide for authors, an abstract that describes activity rather than results, figures unreadable at screen size, English that forces the editor to reconstruct sentences. Language matters here in a specific way: editors penalize ambiguity, not accent. A sentence they cannot parse is a sentence they cannot evaluate.
Design and methods problems do get papers desk rejected, but they have to be visible in the abstract or in a fast skim of the methods: no control condition, an outcome that cannot answer the stated question, a sample so small the analysis cannot support the claim. Subtle problems survive the screen and die in review instead.
A 2024 study in Communications in Transportation Research analyzed over 5,000 rejected transportation manuscripts and found that measurable submission attributes, including text similarity and duplicate submission behavior, tracked with desk rejection outcomes. The finding generalizes past transportation: a meaningful share of screening decisions run on signals that exist in the file you upload, before anyone reads your argument.
Which desk rejection reasons can you actually control?
Nearly all of them, but at different points in the process, which is why authors who “proofread carefully” still get screened out.
Fit is controlled by journal choice and framing, and it has to be settled before you write the cover letter. You cannot fix a scope mismatch with better prose. Compliance is controlled by a checklist you run against the manuscript file, and it is the cheapest category to eliminate completely: there is no reason to lose a submission cycle to a missing ethics approval number. Presentation is controlled by revision passes, and it is the one most authors already attempt.
Novelty framing sits in between. Editors are judging the claim you make, not the work you did, and the same results can be framed as a small extension of a known method or as the first test of that method under a condition where it was expected to fail. The second framing survives screening more often. This is not spin: it is naming which question your paper answers for that journal’s readers.
How do you screen your own paper the way an editor would?
Run the five checks below in order, in about an hour, before you submit anywhere. Stop at the first failure and fix it, because later checks assume the earlier ones passed.
1. The one-sentence result test. Print the title and abstract on a page by themselves. Hand it to a colleague outside your subfield and ask them to say, in one sentence, what you found. Not what you studied. What you found. If they can only describe your topic, your abstract states activity rather than results, and the editor reading it under time pressure will reach the same wall. Rewrite until the finding is in the first two sentences with its magnitude attached.
2. The table-of-contents test. Open the journal’s two most recent issues and list every article title. Then answer two questions. How many of those papers do you cite, and would any of those authors read yours? If you cite none of them, an editor will read that as evidence you are not in conversation with this journal’s community, whatever your reference list looks like otherwise. Also check the article type: many rejections are simply a research article submitted where the journal wanted a brief report, or a 9,000-word manuscript sent to a 4,000-word section.
3. The guide-for-authors diff. Open the journal’s author guidelines and make a literal list of every requirement stated with “must” or “required”: word limits, structured abstract headings, reporting checklist, data availability statement, ethics approval and consent language, trial registration number, conflict of interest disclosure, author contribution statement. Tick each one against the manuscript file, not against your memory of it. If the journal names a reporting guideline, get the current checklist from the EQUATOR Network and complete it honestly, including the items you cannot satisfy. Editors notice a completed checklist with candid gaps; they also notice a checklist that claims compliance the paper does not show.
4. The similarity pass. Run a similarity check yourself, and read the report rather than the score. Reused methods boilerplate from your own earlier papers is the usual source, and it is the kind of overlap that ends a submission at screening without a conversation. Rewrite the overlapping passages or cite the earlier paper explicitly for the protocol.
5. The cover letter test. Write three sentences: what you found, why this journal’s readers specifically need it, and what is new relative to the closest published work. If sentence two comes out as praise for the journal’s reputation, you have no argument for fit, and the screen is where that shows up. Either find the argument or find a different journal.
Some of that is mechanical enough to automate. Compliance items, missing statements, internal inconsistencies between abstract and results, and reference problems are the kind of thing dedicated academic tools, Fukuro among them, can flag in a pass over the manuscript. What no such tool can judge reliably is fit and novelty for a particular journal’s readership, which is where a large share of desk rejections are actually decided. Automated checks reduce the avoidable failures. They do not replace a senior colleague’s read, and they do not stand in for peer review.
What does a desk rejection letter actually tell you?
Usually one usable sentence, buried in boilerplate, and it is worth extracting before you resubmit.
Letters follow a template: thanks for the submission, we receive many more manuscripts than we can publish, we have decided not to send yours for review, we wish you success elsewhere. That paragraph carries no information. The signal is the clause that names a reason, and the common phrasings map to different next actions.
“Does not fall within our scope” means submit elsewhere, unchanged, today. “Not of sufficient priority for our general readership” is an impact judgment: the work is fine, the claim is too narrow for this venue, and a specialist journal is the right move. “We cannot consider manuscripts without institutional ethics approval” is a compliance failure you fix in an afternoon and can resubmit against, sometimes to the same journal. Silence about reasons, with only the template, most often means fit.
One thing a desk rejection almost never tells you is that your science is wrong. Nobody assessed it.
What should you do after a desk rejection?
Resubmit within days. The manuscript carries no mark from a desk rejection, no record follows it, and the fastest authors treat the letter as routing information rather than as feedback.
Before you send it on, do three things. Fix the named reason if there was one. Reframe the title, abstract, and introduction for the new journal’s readership, since the framing that failed a general journal on priority often works verbatim at a specialist one. Then run checks 2 and 3 above against the new target, because the guide for authors changed even if your paper did not.
If the journal offers a transfer or cascade to a sister title, weigh it against the delay it saves. The transfer usually preserves your submission and any editorial notes, which is worth something, but the receiving journal may be a poorer fit than one you would choose yourself. Read the offered journal’s recent issues before accepting.
Appeals are the exception, not the route. COPE guidance treats appeals as a mechanism for procedural or factual errors, not for disagreement with an editorial judgment about scope or priority, and appeal outcomes reflect that. Weeks spent contesting a screening decision are weeks the paper is not under consideration anywhere.
The useful property of desk rejection is that it is reproducible. The criteria are published, the order is stable, and the whole screen fits in an hour of your own time. A paper that fails it fails on things you could have seen, which is a better problem to have than the alternative.