What to Do After a Paper Is Rejected

"Most of what I reject gets published. I just wish I could choose which version of it ends up in print."

A handling editor at a large methods journal said that to me over bad conference coffee years ago. We'd been talking about desk rejections, and she was tired in the way editors get tired in their third year. She wasn't worried that rejected papers disappear. They almost never do. What bothered her was that so many came back out somewhere else with the same unfixed problems, sometimes in a journal that should never have taken them.

I've thought about that line every time I've written a rejection letter since, and every time I've received one. A rejection isn't the end of a paper. It decides which version of the paper you publish next. This post covers how to make that the better version without losing six months to it.

Give it two days, then read the letter like a document

Everyone says "don't respond emotionally." Fine. More practically: don't make any decisions for 48 hours. Don't pick a new journal, don't email the editor, and don't start reformatting. Read the decision once, close it, and go do something else.

When you come back, read it the way you'd read someone else's methods section, looking for information and ignoring tone. Reviewers are often blunt because they're busy, not because they're hostile. A comment like "the analysis is inappropriate" usually means "I couldn't tell why you did this," and that is fixable.

I tell people to copy every comment into a single document and strip out the adjectives. "This deeply flawed sampling approach" becomes "concern about sampling approach." Once the adjectives are gone, most decision letters turn out to be a manageable list.

Work out which kind of rejection you got

"Rejected" covers several different decisions, and each one calls for a different next move.

It's worth separating the editor's letter from the reviews. The editor's paragraph is where the decision was actually made. If it says "the reviewers raise concerns about generalizability," that's the issue that sank you, even if Reviewer 1 wrote three pages about figure labels.

Use the reviews even though nobody is making you

This is where I disagree with a lot of common advice. You'll hear people say, "Different journal, different reviewers, just send it as is." Sometimes that works. Often it doesn't, because fields are small. I've reviewed the same manuscript for two journals in the same year, unchanged, with my earlier comments still unaddressed. I didn't write a kinder report the second time.

Not every comment deserves a change, though. I sort them into four piles:

1. Fix: errors, missing analyses, unclear methods, missing reporting items. Do all of these.

2. Clarify: comments where the reviewer misunderstood something. A misunderstanding is evidence that the text was unclear, so rewrite it.

3. Consider: reasonable suggestions that would change the paper's scope. Decide deliberately whether to take them on.

4. Set aside: requests for a different study, or preferences that conflict with your field's norms. Note them and move on.

The "clarify" pile is where most of the value is. Here's a typical case.

Reviewer comment: "The authors do not justify excluding participants with missing follow-up data, which risks bias."

Before: "Participants with incomplete follow-up were excluded."

After: "Of 412 enrolled participants, 37 (9%) lacked 12-month follow-up. Baseline characteristics did not differ meaningfully between completers and non-completers (Supplementary Table 2). As a sensitivity analysis, we used multiple imputation; estimates were similar (Supplementary Table 3)."

The reviewer may well have been wrong about the bias. The original sentence just gave them no reason to think so.

The cheapest time to fix methods is right now

A few years ago, Maren, a librarian who runs systematic review workshops at a university I worked with, showed me a rejected review that a team had brought to her. The reviewers had flagged that the search couldn't be reproduced. The team had listed databases but gave no full search strings, no dates and no flow diagram. The authors' first instinct was to send it to a lower-tier journal and hope nobody noticed.

Maren sat with them for two afternoons. They rebuilt the search, documented it fully, re-ran it and added a flow diagram that followed the PRISMA 2020 statement. They found eleven new eligible studies. The conclusion didn't flip, but the confidence intervals tightened and the discussion got more honest. The paper went to a better journal than the one they'd been planning to settle for.

What she told them has stuck with me: "A rejection is the only time in a paper's life when you're allowed to fix the foundation without anyone watching." Once the paper is published, every fix becomes a correction notice.

If your study type has a reporting guideline (CONSORT for trials, STROBE for observational studies, PRISMA for reviews, and dozens of others), use the rejection as a reason to complete the checklist properly. The EQUATOR Network indexes them all. Many reviewer comments are simply missing checklist items described in prose.

Choosing the next journal

The default strategy is to move one step down the impact ladder. I think that's lazy, and it's often wrong. A paper rejected for scope from a high-profile general journal may fit better at an equally strong specialty journal. A paper rejected because the contribution felt small might belong somewhere that explicitly values replications or null results.

Here's how I actually choose:

• Read five recent papers from the candidate journal. Read the papers themselves, not the aims and scope page. Ask whether yours would look normal sitting next to them in the table of contents.

• Check where your own reference list publishes. If you cite a journal eight times, its editors and reviewers probably care about your question.

• Look at transfer options. Many publishers offer to pass your manuscript and its reviews to a sister journal. This can save months. The downside is that you're often offered a journal you'd never have chosen. Accept the transfer only if you'd have picked that journal yourself.

• Check turnaround and article processing charges before you spend a week reformatting.

Rejected authors are the target market for predatory journals

Within days of a rejection, your inbox will fill with flattering invitations promising fast decisions. This is not a coincidence. A recently rejected, deadline-pressured researcher is exactly the person those journals are looking for.

I once asked Dev, a grant officer I've worked alongside for years, how he views publication records in progress reports. He said a manuscript listed as "under review" at a credible journal is completely fine. Reviewers of the next application understand that peer review takes time. What worries him is a paper rushed into a questionable outlet to have something "accepted" before a reporting deadline. "That's a line item I have to explain," he said. "Under review never is."

Before submitting anywhere unfamiliar, run it through the Think. Check. Submit. checklist. If the journal is open access, check whether it's indexed in the Directory of Open Access Journals. A missing DOAJ listing doesn't prove anything on its own, but it should send you looking more carefully.

Common mistakes when resubmitting elsewhere

These are the ones I see from the editor's side, roughly in order of how often they happen.

Leftovers from the last journal

The most common one is a cover letter addressed to the previous editor. Close behind are the previous journal's reference style, its section headings, a "Highlights" box the new journal doesn't use, and a word count that's 30% over the limit. None of these get a paper rejected on their own. They do tell the editor that the paper was turned down somewhere and not much was done afterward.

Submitting to two journals at once

Don't. It's duplicate submission, and both the ICMJE recommendations and COPE guidance treat it as misconduct. You can't submit elsewhere until you've received a final decision or formally withdrawn.

Quiet authorship changes

Revisions sometimes bring in a new analyst or a new co-author. That's fine, but every author has to agree to the change in writing, and contributions should be updated. If the journal uses CRediT contributor roles, revisit them properly. Don't just add a name at the end.

Stale literature

If six months or more have passed since your first submission, re-run your search. A reviewer at the new journal may know about the paper that came out in the meantime, and "the authors fail to cite recent work" is an easy criticism to avoid.

Hiding the history when disclosure would help

Some journals ask whether the manuscript was previously reviewed elsewhere, and a few accept prior reviews with your response. If you made substantial improvements, saying so can work in your favor. A line like this helps:

"An earlier version of this manuscript was reviewed at another journal. We have since added a sensitivity analysis addressing missing data, expanded the search to 2026, and revised the framing of our primary outcome. We are happy to share the prior reviews and our responses if useful."

Editors read that as an author who takes review seriously.

A short pre-resubmission checklist

• Every "fix" and "clarify" comment from the last round has been addressed in the text.

• The reporting guideline checklist is complete and matches the manuscript.

• Literature search re-run; new relevant papers cited.

• Title and abstract rewritten for the new journal's readership, not just trimmed.

• Cover letter addressed to the right editor and journal, explaining why this paper fits there.

• Formatting, reference style, word count and figure specifications match the new author guidelines.

• All authors have approved the revised version and the target journal.

• Data availability, funding, conflict of interest and ethics statements updated.

• No previous journal's name anywhere in the files. Search the documents for it.

When to stop revising and rethink

One rejection tells you about a journal. Three rejections for the same reason tell you about the paper. If several editors independently say the contribution is unclear, the problem is probably the framing and not the journals. Try writing the paper's main claim in one sentence for someone outside your subfield. If you can't, the next editor won't be able to either.

Sometimes the honest answer is that the paper is really a smaller, cleaner paper, or a preprint plus a data paper, or a short report. That isn't failure. Do resist the opposite temptation of splitting one study into several thin papers to increase the count. That's salami publication, and editors notice.

A practical note on the logistics. The worst part of resubmitting is usually reassembling everything: the reviews, your triage notes, the updated search, the new draft and the co-author comments, all scattered across email and folders. I've started keeping all of it in one place. PACR is an AI-native research workspace built for this kind of continuity. You can re-run your literature search in its research search and exploration tools, summarize new papers, and carry those notes straight into the revised draft and the cover letter without losing track of which reviewer comment drove which change. If that would save you a week this cycle, you can set up a free workspace before you open the decision letter again.

The editor I had coffee with was right. Your paper will very likely be published somewhere. In the next two weeks, you get to decide which version that will be.