Response to Reviewers Letter Template
The full skeleton plus the four replies that are hard to write: agreement, respectful disagreement, a change you cannot make, and two reviewers who want opposite things.
A response to reviewers is read by two audiences with different questions. The editor wants to know whether the substantive objections were dealt with, in about two minutes. The reviewers want to see their own comment, the change it produced, and where to find it in the file. Almost every structural rule in this template follows from serving both at once. Use it alongside the guide to responding to reviewers, which covers the judgement calls this page only gives you wording for, and run the revised file through the manuscript checker before you resubmit, because revision is where new formatting problems get introduced.
The letter opening
The first half page decides how the rest is read. It thanks the reviewers, states what changed at the level of the paper rather than the sentence, and explains the conventions you used so nobody has to work them out. Keep it to one screen.
Dear Professor [Editor name], Re: Manuscript [ID], "[Title]" Thank you for the opportunity to revise this manuscript, and please pass our thanks to both reviewers. The comments were detailed and the paper is materially better for them. Three substantive changes were made. First, we added the sensitivity analysis requested by Reviewer 1 (new Section 3.4 and Table 4), which does not change the direction of the main result but narrows the range over which it holds. Second, we rewrote the Discussion to separate what the data support from what we speculate, following Reviewer 2's comment on overreach. Third, we added the two missing references and corrected the unit error in Table 2. Below we respond to every comment in turn. Reviewer comments are in italics, our responses in plain text. Page and line numbers refer to the clean revised manuscript; a tracked-changes version is also uploaded. We hope the revision is now suitable for publication in [Journal]. Yours sincerely, [Corresponding author, on behalf of all authors]
- Summarise at the level of the paper
- three substantive changes, each with the section it landed in. An editor deciding whether to send the paper back out reads only this.
- Say what did not change as well as what did, when a headline result survived: that is the sentence the editor is looking for.
- State the conventions once
- how comments are marked, which file the line numbers refer to, and that a tracked version exists.
- Thank the reviewers once, briefly, at the top. Repeating gratitude before every single response makes a long letter longer without adding anything.
- Do not claim in the summary that you addressed every comment if you declined some. The editor will reach the declined ones anyway, and finding them after that claim is the worst order to read them in.
Reply 1: a comment you accept
The majority of comments fall here, and the only failure mode is vagueness. A response that says the text has been clarified forces the reviewer to hunt for the change. Quote the comment, name the change, give the location, and paste the new text when it is short enough to read in place.
Reviewer 1, comment 3: "The sample size calculation is not reported, so it is impossible to judge whether the study was adequately powered." Response: We agree, and this was an omission. We have added the a priori power calculation to the Methods (page 7, lines 145 to 152) and stated the assumptions it rests on. Added text: "Sample size was calculated a priori for a two-sided test at alpha = 0.05 with 80% power, assuming an effect size of d = 0.5 based on the pilot data reported in [12]. This required 64 participants per arm; 71 were recruited per arm to allow for attrition." We also note in the Limitations (page 19, lines 402 to 405) that the assumed effect size came from a pilot with a small sample, and that the study is underpowered for the secondary outcomes.
- Quote the comment verbatim rather than paraphrasing it
- a paraphrase can look like an attempt to answer an easier question.
- Give page and line numbers from the clean file, and say in the opening which file they refer to.
- Paste the new text when it is a paragraph or less. Reviewers read this letter, not the manuscript, on the first pass.
- Add the consequence you noticed while making the change, as with the limitation above. It is the cheapest way to show the revision was thought about rather than executed.
- Say plainly that the reviewer was right when they were. It costs nothing and it sets the tone for the responses where you disagree.
Reply 2: a comment you disagree with
Respectful disagreement is legitimate and editors expect it. What makes it work is that the reasoning is evidence-based and that you still move: acknowledge the concern, give the argument, and where possible make a smaller change that addresses what prompted the comment even though you did not do what was asked.
Reviewer 2, comment 5: "The authors should use a mixed-effects model rather than the repeated-measures ANOVA reported here." Response: We appreciate this suggestion and considered it carefully, but we have retained the repeated-measures ANOVA, for two reasons. First, the design is fully balanced with no missing observations at any time point, the condition under which the two approaches are known to give equivalent fixed-effect estimates; the advantage of a mixed model is in handling unbalanced or missing data, which does not arise here. Second, the assumptions of the ANOVA were tested and met: Mauchly's test was non-significant (W = 0.94, p = 0.31), so no correction was required. To make this reasoning visible to readers rather than only to the reviewer, we have added a sentence to the Statistical Analysis section (page 9, lines 188 to 191) stating why the ANOVA was appropriate given the balanced design. We also ran the mixed-effects model as a check. The estimates agree with the reported analysis to two decimal places, and we are glad to add this as a supplementary table if the editor considers it useful.
- Open by acknowledging the comment before you disagree, and never with a bare statement that you disagree.
- Give a reason a third party can check
- an assumption test, a design feature, a citation. An appeal to your own judgement is the weakest available argument.
- Do something anyway when you can
- a clarifying sentence, a supplementary analysis, a limitation. A response that changes nothing at all is the one most likely to come back.
- Offer the extra analysis rather than attaching it, when it is large. That leaves the decision with the editor, which is where it belongs.
- Keep the register flat. Sarcasm and wounded pride survive into the editor's inbox exactly as written.
Reply 3: a change you cannot make
Some requests are reasonable and simply out of reach: an experiment that needs an animal licence you do not hold, a cohort that no longer exists, a follow-up that would take two years. Treating these as disagreements is a mistake. Say what the obstacle is, say what you did instead, and let the editor weigh it.
Reviewer 1, comment 7: "The mechanism should be confirmed in a knockout mouse model." Response: We agree that a knockout model would be the definitive test of the mechanism we propose, and we would like to run it. It is not achievable within this revision: the line would have to be generated and back-crossed, and our institutional animal licence covers the current protocol only, so the work would take an estimated 18 months and a separate ethics application. We have made three changes in response to the underlying concern. We softened the mechanistic claim in the Abstract and Discussion, which now state that the data are consistent with the proposed mechanism rather than that they establish it (pages 2 and 17). We added the pharmacological inhibition experiment as an independent line of evidence (new Figure 5), which supports the same interpretation through a different route. And we state explicitly in the Limitations that genetic confirmation is outstanding and is the subject of ongoing work (page 19, lines 410 to 414). If the editor considers genetic confirmation essential for publication in [Journal], we understand, and would rather be told now than after a second round.
- Name the concrete obstacle
- a licence, a closed cohort, a timeline. An unexplained refusal reads as unwillingness.
- Answer the concern behind the request. A reviewer asking for a knockout is asking you to stop overclaiming, and weakening the claim addresses most of that.
- Add whatever independent evidence you do have, even if it is weaker than what was asked for.
- Put the gap in the Limitations, in the paper, not only in the letter. That is what the reviewer is checking on the second pass.
- Invite the editor to overrule you. A clean decision now is better than a second round that ends the same way.
Reply 4: two reviewers who want opposite things
It happens often enough to plan for: one reviewer wants the theory section expanded, the other wants it cut. You cannot satisfy both, and quietly siding with one leaves the other reading a revision that ignored them. Name the conflict, choose, and give the editor the reasoning so they can overrule you.
Reviewer 1, comment 2: "The theoretical framing in Section 2 is too compressed to follow and should be expanded." Reviewer 2, comment 1: "Section 2 is long relative to its contribution and could be cut substantially." Response: We note that the reviewers take opposite views of the same section, so we have tried to serve the concern common to both, which we read as the section not earning its length. We have restructured rather than only lengthened or shortened it. Section 2 is now half a page shorter overall (page 4), with the historical background moved to Supplementary Note 1, and the derivation that Reviewer 1 found compressed expanded from three lines to a full paragraph with the intermediate step stated (page 5, lines 96 to 108). The net effect is a shorter section in which the part that carries the argument is longer. We are of course happy to go further in either direction if the editor has a preference.
- State the contradiction openly, quoting both comments together. Editors are used to it and it explains immediately why one reviewer did not get what they asked for.
- Look for the concern the two comments share, which is often a third thing neither of them said.
- Explain the trade-off in one sentence, then hand the choice to the editor. They can see both reports and you cannot.
- Answer both reviewers in the same place, and cross-reference it from the other reviewer's numbered section so neither has to find it.
- Never resolve it by claiming one reviewer misunderstood the paper unless you can show precisely where, because the editor will check.
Frequently asked questions
Follow the submission system. Most journals provide a dedicated upload for the response to reviewers and a separate cover letter field; when only a cover letter field exists, the response goes there. Never put substantive responses only in the manuscript and assume the reviewer will find them.