A revise-and-resubmit decision is good news: it means the editor considers your paper worth further work. Yet many authors stall at exactly this point, because a long list of reviewer comments feels like a verdict rather than a work plan. It is not. Reviewer comments are a checklist for making the paper stronger, and responding to them well is a skill with a clear method: read everything first, decide on each comment, then respond point by point with evidence the reviewer can verify.
This guide walks through that method step by step: how to read the reports without reacting too quickly, how to sort comments into decisions, how to write the response letter, how to agree and how to disagree politely, and how to handle the difficult cases — contradictory reviewers, requests you cannot meet, and comments that feel unfair. For background on how review works, see our explainer on double-blind peer review explained.
Read the editor's decision letter and every reviewer report in full before you change a single word. The editor's letter matters most: it tells you which comments the editor considers essential, which are optional, and what decision the revision must support. Reviewers may disagree with each other, and the editor's letter usually signals how to weigh those disagreements.
Then put the reports aside for a day if you can. First reactions to criticism are rarely your best thinking, and a response written in frustration shows in its tone even when the words are polite. When you return, read the reports again with a pen in hand and mark each comment as one of three kinds: a factual correction, a request for clarification or extra analysis, or a judgement about the paper's argument. This sorting turns an overwhelming list into a work plan.
Discuss the reports with your co-authors before you start revising. Authors often disagree about which comments to accept, and those disagreements are far easier to resolve before the revision than after one author has already rewritten a section. Agree on the approach for each major comment, and record who will handle each change.
For each comment, make one of three decisions: accept and change, accept in part and change differently, or respectfully decline with reasons. There is no fourth option. Ignoring a comment — leaving it unanswered in the hope the reviewer will not notice — is the one response that reliably damages a revision, because reviewers do check, and an unanswered comment reads as carelessness.
Accept most comments. Reviewers are usually right about clarity problems, missing detail, and errors, even when their suggested fix is not the one you choose. When a reviewer misunderstands a passage, the passage usually needs rewriting rather than a rebuttal: if one careful reader misunderstood it, other readers will too. Treat misunderstanding as a signal about your writing, not as a failure of the reviewer.
Decline only with evidence. A respectful disagreement needs a reason grounded in data, literature, or method — not preference. "We prefer our original wording" is not a reason. "We kept the original analysis because the suggested alternative assumes a condition our study design does not meet, as shown in the cited reference" is a reason an editor can accept. If you cannot produce such a reason, accept the comment.
The response letter is as important as the revised manuscript. Its job is to let the editor and reviewers verify every change quickly, without hunting through the paper. The most reliable format is a point-by-point reply, and journals often state the format they want — follow it exactly if they do.
Open by thanking the editor and reviewers for their time and comments, and by summarising the revision in two or three sentences: what you changed overall, and any major decision the editor should know about. Keep it genuine and brief. Thank reviewers for specific helpful points where you can; a generic paragraph of praise that could fit any paper reads as insincere.
Work through the reports in order: the editor's comments first, then each reviewer in turn. For every comment, quote or closely paraphrase it, state what you did, and give the location of the change by section and, where possible, page or line number. A reply that says "we revised the methods" without saying where is almost as unhelpful as no reply. A reply that says "we added the sample selection criteria and final sample size to the Methods section" lets the reviewer verify the change in under a minute.
Match the tone to the action. For changes you made, describe them plainly. For changes you made differently from what was suggested, say what you did instead and why. For comments you declined, explain the reason respectfully and cite the evidence. Keep each reply focused on that one comment; a long reply that wanders into other issues makes verification harder.
Close by listing any changes that affect the whole paper — a revised title, a restructured section, updated numbers — and by confirming that all authors have approved the revised version. If you changed author details, funding statements, or ethics declarations during revision, say so explicitly; journals handle such changes formally, and a quiet change discovered later causes difficulty.
Agreeing is straightforward: say what you changed and where. The skill lies in disagreeing well. A good disagreement has four parts: acknowledgement of the point, a clear statement that you have not made the suggested change, the evidence-based reason, and what you did instead. For example: "We appreciate this suggestion. We have not added the proposed analysis because our sample was not designed for that comparison, and running it now would risk a misleading result. Instead, we have added a paragraph to the Discussion noting this as a direction for future work." That response respects the reviewer, explains the decision, and improves the paper anyway.
What to avoid: defensive language, attributing motives to the reviewer, and dismissing a comment without engaging with it. Phrases such as "the reviewer clearly did not read the paper" or "this comment is wrong" damage your standing with the editor even if the underlying point is fair. If a comment is based on a misunderstanding, correct the misunderstanding by improving the passage and explaining the clarification — the reviewer will usually accept that graciously.
Be careful with partial agreement. If you accept the spirit of a comment but implement it differently, say so explicitly rather than writing as if you followed the suggestion exactly. Reviewers compare the letter with the manuscript, and a claim of agreement that the manuscript does not support reads as dishonesty. Precision in the letter builds trust; vagueness destroys it.
Contradictory reviewers are common: one asks you to shorten the discussion, another asks for more detail. Do not try to satisfy both fully. State the contradiction plainly in your response, explain the choice you made and why, and let the editor adjudicate. Editors expect this and would rather see an explicit choice than a paper that tries to do two incompatible things at once.
When a reviewer asks for work you cannot do — an additional experiment, data you do not have, an analysis your design cannot support — explain the constraint honestly, offer the closest alternative you can provide, and discuss the limitation in the paper. Reviewers and editors generally accept a reasoned explanation; what they do not accept is pretending the gap does not exist or claiming work that was not done.
If a comment feels unfair or mistaken, test it before rejecting it. Ask a colleague to read the comment and the relevant passage without telling them your view. Fresh eyes often reveal that a comment you found unreasonable points to a genuine weakness in wording or evidence. If the comment is still wrong after that check, disagree with evidence, briefly and courteously, and move on. One well-argued disagreement strengthens a response letter; five weak ones weaken it.
Finally, keep the revision deadline in view. If you need more time than the journal allowed, ask the editor before the deadline passes, with a brief reason and a realistic new date. Editors routinely grant reasonable extensions when asked in advance; a silent missed deadline is much harder to repair.
Make the changes the letter promises, exactly as promised, and mark them in the way the journal asks — usually tracked changes, a highlighted version, or both, plus a clean version. Before resubmitting, read the whole revised manuscript again from start to finish. Revisions introduce their own errors: repeated paragraphs, numbers that no longer match the tables, references added to the text but not the list, and a conclusion that no longer matches the revised results. A clean revised manuscript is part of what the reviewer is judging.
Check consistency carefully: every number in the revised abstract should match the main text, every new citation should appear in the reference list, and every new table or figure should be numbered and cited in the text. Run the journal's own checks again — word limits, reference style, file formats — because a revision that fails a formatting check can be returned before it reaches the reviewers. See our guide to common rejection reasons for the checks editors apply at screening.
Do not use the revision to rewrite the paper into a different paper. A revision should strengthen the submitted argument, not replace it with a new study. If the reviewers' comments reveal that the paper needs a fundamentally different design or a new dataset, that is usually a matter for a new submission, and the honest response is to say so to the editor rather than forcing the change.
A revision done this way usually takes longer than authors expect, and that is normal. The time is well spent: a response letter that lets reviewers verify changes quickly is one of the strongest signals you can send an editor, and it often makes the difference between a second revision and an acceptance. If your paper was rejected rather than returned for revision, our guide on how to publish a research paper explains how to revise and submit elsewhere.
| Comment type | Response strategy | What to write in the letter |
|---|---|---|
| Factual correction | Fix it and say where | "Corrected in the Methods section, with the revised figure." |
| Request for clarification | Rewrite the passage, then explain the change | "We have rewritten this paragraph to state the criteria explicitly." |
| Request for extra analysis | Do it if possible; if not, explain and offer an alternative | "We have added the analysis to the Results section" or a reasoned explanation of the constraint. |
| Judgement you agree with | Accept and revise accordingly | "We agree and have restructured the Discussion as suggested." |
| Judgement you disagree with | Decline politely, with evidence, and state what you did instead | "We have not made this change because... Instead, we have..." |
| Contradictory reviewers | State the contradiction, choose, and let the editor decide | "The reviewers differ on this point. We have chosen... for the following reason." |
These authoritative external resources explain the standards and guidance mentioned in this article:
Do I have to accept every reviewer comment?
No, but you must answer every comment. Accept most comments, and decline only with an evidence-based reason, stated politely, plus what you did instead.
What if two reviewers contradict each other?
State the contradiction in your response letter, explain the choice you made and why, and let the editor adjudicate. Do not try to satisfy both fully.
How long should a response letter be?
As long as it needs to be to answer every comment point by point, with the location of each change. A thorough letter is a strength, not a weakness.
What if a reviewer asks for work I cannot do?
Explain the constraint honestly, offer the closest alternative you can provide, and discuss the limitation in the paper. Reviewers generally accept a reasoned explanation.
Should I mention the reviewers by name in my response?
Only if the review model is open and the journal asks for it. In double-blind review, refer to "Reviewer 1" and "Reviewer 2" as the journal's reports label them.
Can I ask the editor for more time to revise?
Yes. Ask before the deadline passes, with a brief reason and a realistic new date. Editors routinely grant reasonable extensions requested in advance.