The response letter is a document with a job. Most of them do that job badly, and it costs authors a revision cycle they did not need to spend.
An editor reading your revision is answering one question: has this manuscript become publishable. They are not re-reviewing it from scratch. They are checking whether the specific objections raised last time have been dealt with, and they are usually doing that with the reviewer reports open in another window.
That tells you the shape of the document. It is a checklist the editor can run down, not an essay about your revision. Every comment gets an entry. Every entry says what you did and where to find it. An editor who has to hunt through the manuscript to verify a claim you made about it will send the paper back out for another round.
The most common failure is not rudeness or defensiveness. It is a letter that describes changes in general terms. "We have clarified the methods section" forces the editor to go and check. "We have added the power calculation to Methods, paragraph 3, page 8" does not.
Number every comment. Quote it verbatim. Answer underneath it. Point at the change.
Reviewer 2, comment 4
"The authors report a significant association but do not account for multiple comparisons across the 47 biomarkers tested."
We agree. We have applied Benjamini-Hochberg correction at a 5% false discovery rate across all 47 comparisons. Twelve biomarkers remain significant after correction. Methods, page 9, paragraph 2. Table 3 now reports both raw and adjusted p-values. The Discussion has been revised accordingly, page 17, paragraph 1.
Quoting the comment matters more than it looks. Reviewers read the response letter to check whether you understood them. Paraphrasing an objection is where authors quietly answer an easier question than the one they were asked, and reviewers notice.
Give page and paragraph, not just section names. A reviewer who can verify your claim in ten seconds will accept it. One who cannot will reread the section.
Every comment resolves to one of three answers. Deciding which one you are giving, before you start writing, prevents most bad response letters.
You agree and you changed it. The easy case. Say what you changed and where. Do not explain at length why the reviewer was right. Agreement needs no argument, and padding it makes the letter longer without making it more convincing.
You disagree. Harder, and covered below. The short version is that you must give a reason a third party could evaluate, and you must still change something.
You agree but cannot do it. The reviewer is right that a longitudinal design would answer the question better, and you have cross-sectional data that took four years to collect. Say so plainly, then convert the objection into a stated limitation. A reviewer who sees their point acknowledged in the Limitations section will usually let it go. One who sees it silently dropped will not.
Authors are told to be diplomatic, which is true and not very useful. The mechanism that actually works is narrower: move the disagreement onto ground the editor can adjudicate.
"We respectfully disagree with the reviewer's characterisation" is not adjudicable. It is your opinion against theirs, and the editor has no way in. A citation, a reanalysis, or a sentence from the manuscript that the reviewer appears to have missed are all adjudicable. The editor can check them.
Concede the part that is right before you contest the part that is not. Reviewers are frequently correct about the observation and wrong about the implication. Separating those two things gives you somewhere to agree, which makes the disagreement read as considered rather than reflexive.
And change something anyway. If a trained reader misread your argument, the argument can be stated more clearly, whatever you think of their conclusion. A disagreement paired with a clarifying edit almost always sticks. A flat refusal with no change attached is the response most likely to come back.
One reviewer wants the theory section expanded. The other says it is already too long. This happens often, and authors treat it as a crisis when it is routine.
Do not try to satisfy both. You cannot, and attempting it produces a section that is longer and less focused, which fails both reviews at once. Name the conflict explicitly in the letter, state which direction you took, and give the reason. Then flag it for the editor, because resolving reviewer disagreement is the editor's job and they are entitled to make the call.
Reviewer 1 asks us to expand the theoretical framing, while Reviewer 3 considers the current framing too long relative to the empirical contribution. We have followed Reviewer 3 and cut the framing by roughly a third, because the paper's contribution is empirical and the framing was reviewing literature rather than motivating our design. We are happy to restore the material if the editor prefers Reviewer 1's reading.
That paragraph does three things. It shows you read both reviews. It gives a reason tied to the paper rather than to convenience. And it hands the decision to the person whose decision it is.
The expensive mistake is answering the comment you wish you had received.
A reviewer writes that the control group is not comparable on baseline severity. The authors respond with a paragraph about how they matched on age and sex. Both statements are true. They are not about the same thing. The reviewer reads the response, sees that their actual objection is unaddressed, and repeats it in round two with less patience.
This is rarely dishonest. It happens because authors read a comment, recognise the general area it belongs to, and answer from that area. The defence is mechanical: after drafting each response, reread the quoted comment and ask what specific claim it makes. Then check that your paragraph addresses that claim rather than its neighbourhood.
Reviewers repeat unaddressed comments. A revision that misses three objections in round one usually gets them back in round two, and what could have been one revision cycle becomes two. In most fields that is several months.
Thank the reviewers once, at the top. Repeating it before every response is padding, and by comment twelve it reads as nervous rather than gracious.
Write as though a third person will read the exchange, because one will. The editor sees your letter and the reviewer's original comment side by side. That framing removes most of the temptation to score points.
Do not describe a reviewer's comment as unfair, careless, or based on a misunderstanding of the field. If the comment is wrong, demonstrate it with evidence and let the editor draw the conclusion.
Two documents go back with a revision, and they have different readers. The response letter is written for the reviewers, who will see it. The cover letter is written for the editor, who decides.
Authors often merge them, which produces a cover letter that restates every response at length. The editor already has the response letter. What they want from the cover letter is shorter: what changed at the level of the paper, whether any reviewer request was declined and why, and anything they need to rule on.
Four or five sentences is usually right. If a reviewer asked for an analysis you could not run, the editor should learn that from the cover letter rather than discovering it on page eleven of the response. Editors dislike surprises late in a document more than they dislike a declined request.
That last one catches more problems than it should. Letters are often written alongside the edits rather than after them, and a change described in the letter sometimes never lands in the manuscript. Read the letter with the revised paper open and verify each claim against the text.
What to check across methodology, statistics, writing, and formatting before you submit.
The reasons editors decline manuscripts before review, and how to avoid each one.
The errors reviewers catch most often, from misused tests to overstated effects.
How the process works, who decides what, and what reviewers are actually asked to do.
The Reviewer 2 Generator reads an abstract free. The free assessment reads a whole manuscript and tells you where it stands. For a full review, PeerGenius runs a panel of specialist reviewers over the paper and returns their objections before an editor sees it.
PeerGenius is built by Neon Peach, LLC, a product studio.
Run the panel over your manuscript and get the objections before an editor does.
Get free assessment