Desk Rejection Checker
Paste your abstract or introduction to see the language problems that make editors reject papers before review, with what to fix. It checks language only, in seconds, and nothing you paste is stored.
What this checks, and what it doesn't
The checker applies fixed rules (no AI) to the language and reporting signals editors notice first: informal or emotive wording, overclaiming and causal claims from correlations, missing sample sizes, incomplete statistics, common grammar slips, passive voice, wordy phrasing, long sentences and missing abstract structure. Editors also desk-reject papers for reasons no language check can see: a poor fit with the journal's scope, limited novelty, or weak methods. Treat the score as one part of your pre-submission review, not a prediction.
How the Desk Rejection Checker works
What is desk rejection?
A desk rejection happens when a journal editor turns a manuscript down before it is sent to peer reviewers. At many journals, most submissions never reach review. Editors decide quickly, often from the title, abstract and first page, so a paper that reads poorly can be rejected before anyone judges the research itself.
The most common reasons are a poor fit with the journal, a contribution that is not clear, missing parts of the expected structure, and writing that is hard to follow. This checker looks at the last two: the language signals an editor notices in the first minute.
What the checker looks at
Paste at least 50 words of your abstract or introduction. The checker applies fixed language rules (it does not use AI) to the signals editors notice in the first minute and adds them into a score out of 100, where higher means more problems:
Sentence length. Average sentences over about 28 words are hard to read; over 35 is a strong warning sign. Passive voice. When most sentences use constructions like “was conducted” or “were analysed”, the text reads slowly and hides who did what. Wordy or non-native phrasing. Phrases like “in order to”, “it is worth noting that” and “the present paper” add words without meaning. Hedging. Too many words like “possibly”, “somewhat” or “might be” weaken your findings. Structure. An abstract with no sign of background, methods, results or conclusion may look incomplete.
Tone and claims. Informal words (“kids”, “a bunch of”), contractions, dramatic words (“shocking”, “huge”), personal opinion (“my point”), and claims the data can’t support: “proves”, or causal words like “causes” and “ruins” when the study is a survey or correlation. Reporting. A study described with no sample size, statistics without a p-value or confidence interval, LaTeX code pasted into the text, and recommendations that don’t follow from the results. Grammar. Comma splices such as “…, moreover, …” and common agreement slips like “freshman are” or “less students”.
How to use your result
Treat the score as a quick screen, not a verdict. A low score means the language is unlikely to be the problem; a high one means an editor may struggle with your first page. Fix the flagged items, paste the revised text and check again. The checker reads only language: it cannot judge whether your research suits a particular journal.
Your text is analysed on our server to produce the score and is not stored. If you enter your email address we save that address and the score, and email you the breakdown.
Frequently asked questions
Is the Desk Rejection Checker free?
Yes. You see your score without signing up. The detailed breakdown is free too; we ask for an email address and send you a copy.
Can this tool predict whether my paper will be accepted?
No. It measures language and structure signals that editors notice early. Acceptance also depends on the research question, methods, results and fit with the journal, which only expert reviewers can judge.
How much text should I paste?
At least 50 words. Your abstract, or the first two paragraphs of your introduction, gives the most useful result, because that is what editors read first.
How do I lower my score?
Replace informal and dramatic words with neutral ones, report your sample size and full statistics, say “is associated with” unless you tested cause and effect, split long sentences, turn passive sentences active where the actor matters, cut filler phrases and keep hedging for claims that really are uncertain. If English is not your first language, a human copyedit is the fastest way to fix all of these at once.