To respond to reviewer comments, write a point-by-point response letter that quotes each comment, answers it, and says exactly what you changed in the manuscript and where, alongside a revised manuscript with the changes marked. The editor should be able to check every reply in under a minute. That is the whole method; the rest is doing it carefully and in the right tone.
This guide covers what the editor’s decision actually means, how to sort the comments before you write, the structure of a response letter with a template you can adapt, worked examples of difficult replies, and the checks to make before you resubmit. Last reviewed September 2026.
What do minor revision, major revision and reject-and-resubmit mean?
Read the editor’s letter before the reviews. The editor’s summary tells you which comments matter most, and the decision tells you how much risk you are carrying.
- Minor revision. The paper is close. The editor may accept it without sending it back to reviewers. Do everything asked, quickly and precisely.
- Major revision. The editor sees a publishable paper if the concerns are answered. It will usually go back to the same reviewers. This is the most common decision for papers that are eventually accepted in indexed journals.
- Reject and resubmit (or “reject with encouragement to resubmit”). The changes needed are too large for a revision, but the journal would look at a substantially new version as a new submission. It may go to new reviewers.
- Reject. The journal will not consider this paper again. Use the reviews to improve it before you send it elsewhere. If the paper came back with no reviews at all, read about desk rejection first.
Put the deadline in your calendar the day the letter arrives. Revision deadlines vary by journal and decision type. If you will need more time because of new experiments or fieldwork, write to the editorial office well before the deadline, propose a new date and give a one-line reason.
How should you read and sort reviewer comments?
Read the reviews once, then put them away for a day. Almost everyone reads harsh comments as personal on the first pass. On the second reading, copy every comment into a table, split long paragraphs into separate points, and number them (R1.1, R1.2, R2.1 and so on). Then sort each one:
| Type of comment | Example | What to do |
|---|---|---|
| Must-fix | “The regression assumptions are not reported.” | Make the change fully. Say exactly what you added and where. |
| Clarification | “It is unclear how the sample was drawn.” | The reviewer misread because the text allowed it. Rewrite the passage, even if the information was technically there. |
| Reasonable but costly | “A second dataset would strengthen the findings.” | Do it if you can. If not, explain why, and offer a smaller change: a sensitivity check, a limitation, a softer claim. |
| Out of scope | “The authors should also study rural respondents.” | Thank the reviewer, explain why it falls outside the paper’s question, and add it to future work. |
| Mistaken | “The authors did not control for age.” | Correct it politely with the page and line where you did. Then ask why the reviewer missed it and make it easier to find. |
| Language and presentation | “The English needs improvement.” | Have the whole paper edited, not just the sentences they quoted, and say so. |
Two habits help at this stage. First, assume the reviewer is a careful reader who misunderstood because your text let them; that assumption is usually right, and it produces better revisions. Second, discuss the sorted table with your supervisor and co-authors before anyone starts rewriting, so you agree on which requests to accept and which to decline.
How do you structure a response letter?
Editors and reviewers read response letters side by side with the manuscript. A clear, predictable layout saves them time, and a reader who saves time is a friendlier reader. Use this outline:
- Opening paragraph to the editor. Thank the editor and reviewers in one or two sentences. Summarise the main changes in three to five bullet points: new analysis, restructured sections, shortened text.
- Response to Reviewer 1. Each comment quoted in full, in italics or a shaded box, followed by your reply in plain text, followed by the change, with location.
- Response to Reviewer 2, and so on, in the same format.
- Other changes. Anything you changed that no reviewer asked for, such as corrected typos in a table or an updated reference. Editors dislike surprises.
A template you can adapt
Dear Dr [Editor’s surname],
Thank you for the opportunity to revise our manuscript “[Title]” (Manuscript ID [number]). We are grateful to the reviewers for their careful reading. The main changes are:
- [Change 1, for example: added a sensitivity analysis using an alternative measure of X (Section 4.3).]
- [Change 2]
- [Change 3]
Our point-by-point responses follow. Reviewer comments are in italics; page and line numbers refer to the revised manuscript with tracked changes.
Reviewer 1
Comment 1.1: [quote the comment in full]
Response: [Your reply, in two to five sentences.]
Change: [What changed, with page and line numbers, or “No change; see explanation above.”]
…
Yours sincerely,
[Corresponding author, on behalf of all authors]
Keep each reply short. A reply longer than the change it describes is usually defending rather than revising. If a new paragraph is short, paste it into the letter so the reviewer need not look for it. The same plain, courteous tone you used in your cover letter works here.
How do you reply to comments you disagree with?
You can disagree. Reviewers are sometimes wrong, and editors expect authors to say so when they are. What editors do not accept is disagreement without reasons, or a reply that argues with the reviewer instead of with the evidence. These phrases work in most situations:
| Situation | Wording that works |
|---|---|
| Accepting a change | “We agree and have revised the paragraph accordingly (page 6, lines 12–20).” |
| Clarifying a misreading | “We have rewritten this passage to make the sampling procedure clearer (page 4, lines 3–11).” |
| Adding analysis | “As suggested, we ran the model with firm size as a control. The main coefficient remains significant (new Table 5, page 11).” |
| Declining, with reasons | “We considered this carefully. Because the study’s objective is X, adding Y would change the research question. We now state this boundary in the limitations (page 18, lines 2–7).” |
| When reviewers disagree | “Reviewer 1 suggested A and Reviewer 3 suggested B. We followed A because … and have noted B as an alternative (page 9).” |
| Admitting an error | “We thank the reviewer for spotting this. The value in Table 2 was a transcription error and has been corrected; the conclusions are unchanged.” |
Worked example: a reviewer who misread the method
Reviewer 2, comment 3: “The authors have not addressed common method bias, which is a serious concern with self-reported data.”
Weak reply: “We already addressed this in the paper. Please read Section 3.4.”
Better reply: “Thank you. We did run Harman’s single-factor test (the first factor explained 28% of variance), but we agree the discussion was brief and easy to miss. We have moved it into its own subsection, added a marker-variable test, and discussed the remaining limitation (Section 3.5, page 9, lines 4–22).”
The better reply keeps the fact that you were right, but it treats the misreading as a signal to improve the text, and it adds something. That is what gets a reviewer to write “addressed” in the second round.
Worked example: a request you cannot meet
Reviewer 1, comment 5: “Longitudinal data would be needed to support causal claims.”
Reply: “We agree that our cross-sectional design cannot establish causality. Collecting longitudinal data is not possible within this study, so we have removed causal language throughout (for example, ‘leads to’ is now ‘is associated with’, pages 2, 12 and 17) and added this as the first limitation (page 18, lines 1–6).”
You gave the reviewer what the comment was really about, which was overclaiming, without pretending you could run a new study.
What should you check before resubmitting?
Most second rounds are caused by small misses, not big disagreements. Work through this list with the final files open:
- Every comment from every reviewer, and from the editor, has its own numbered reply. Comments hidden inside long paragraphs have been split out.
- Every “we have added” or “we have revised” matches a real change at the page and line you cite, in the final file.
- Line numbers were rechecked after the last edit.
- The abstract, tables, figures and conclusion agree with any new numbers.
- New references exist, are cited correctly and actually support the sentence they sit in.
- The tracked-changes file and the clean file are the same text.
- All co-authors and your supervisor have approved both the manuscript and the response.
- The response contains no sarcasm, no “as clearly stated”, and no comment on the reviewer’s competence.
Resubmit through the journal’s system from the corresponding author’s account, uploading the response letter, the tracked-changes manuscript and the clean manuscript as the system asks. Then expect another wait: a revision that goes back to reviewers usually takes weeks to a few months. Realistic ranges for each stage are in our publication timeline guide.
If you would like a specialist to work through a revision with you, that is what our reviewer response support does.
Sources
FAQ
Questions scholars ask
How long should a response letter be?
As long as the comments require and no longer. For a major revision with three reviewers, 5–15 pages is common, mostly because the comments are quoted in full. Individual replies should be a few sentences each.
Should I address the reviewer as “Dear Reviewer” in each reply?
No. Address the letter to the editor, then use headings for each reviewer. Thank each reviewer once at the start of their section, not before every comment; repeated thanks reads as filler.
A reviewer asked me to cite their own papers. What should I do?
Cite them if they are relevant and improve the paper. If they are not, say politely that they fall outside the paper’s scope. If the pressure seems excessive, you may raise it confidentially with the editor; COPE’s guidance treats coercive citation requests as inappropriate.
Can I submit the revised paper to a different journal instead?
You can withdraw and go elsewhere, but after a revise decision it is usually a poor trade: the new journal starts from zero and often raises the same concerns. Withdraw only if the requested changes would make it a different study.
Can I use an AI tool to draft replies to reviewers?
Check the journal’s policy first. Many publishers allow language help but hold you responsible for every claim, and forbid uploading confidential review material to external tools. Whatever you use, every reply must describe what you actually did, and you must be able to defend it.
