How to respond to reviewer comments: a point by point guide

A decision letter that says "major revision" can feel like a rejection for the first hour. It is not one. It means the editor and the reviewers see enough in your paper to ask you for more work. How you answer decides whether the second round goes well.

Before you write anything

Read all the reports once, then close them and wait a day. The first reaction to criticism is emotional, and a response drafted in that state usually sounds defensive. When you return, read as a colleague would and ask one question about each comment: what is this reviewer actually worried about?

Then number every comment, for example 1.1, 1.2 for the first reviewer and 2.1, 2.2 for the second, and sort each one into one of four kinds:

• Fair and fixable. The reviewer is right, and the paper needs a change.
• A misunderstanding. The reviewer misread something. Usually that means your text was unclear.
• Out of scope. A reasonable idea, but beyond what this paper claims to do.
• Not possible. It needs data, time, or resources you do not have.

If two reviewers raise the same concern in different words, treat it as the real problem with the paper, even if neither of them says it is important.

The structure of the letter

Open with a short note to the editor: thank them, and summarize the main changes in three to five lines. Then go reviewer by reviewer, comment by comment, in the original order. For each comment, repeat it in full, give your response, and say what changed and where.

• Answer every comment. Including the small ones: a typo, a missing reference, a figure that is hard to read. Skipping any comment tells the reviewer you did not read it.
• Be specific about location. Give the section, page, or line number in the revised manuscript, so the reviewer does not have to hunt.
• Make the changes visible. Send a revised version with changes highlighted, and a clean version if the journal asks for one.

Here is an illustrative example of one comment, written the way I recommend.

Notice what the response does. It agrees with the substance, says what was done, and points to the place where the reviewer can check it.

How to agree

Thank the reviewer once, briefly, and move on to the work. Praising every comment as "an excellent point" starts to sound hollow by the third page. What persuades a reviewer is a concrete action, not an adjective.

How to disagree without losing the editor

You are allowed to disagree. A revision where the authors changed everything the reviewers asked is not always a better paper. But disagreement has to be earned with evidence and delivered without heat.

1. Acknowledge the concern behind the comment. Show that you understood why the reviewer raised it.
2. State your position and your reason. One or two sentences.
3. Show evidence. A result, an additional analysis, a citation.
4. Change the paper anyway. If one careful reader had this objection, others will too. Add a sentence or a paragraph so the next reader does not make the same assumption.

A phrase that works: "We understand the concern that the sample may not represent other settings. We kept the analysis as it is because the study question concerns this population, and we have added a paragraph to the limitations section to state this clearly."

Avoid words that attack the reviewer: "obviously", "the reviewer failed to", "it is clear that the reviewer did not read". Also avoid disagreeing with everything. Concede what is fair, so that your disagreement on a few points carries weight.

When the reviewer asks for something you cannot do

If a new experiment is feasible, do it. If it truly is not, for example because the data does not exist or the compute would take months, say so plainly. Explain what you did instead, such as a smaller analysis or a discussion of the issue, and add the point to the limitations. An honest limitation is a better answer than a hurried experiment that you cannot defend.

When the reviewer misunderstood

Treat a misunderstanding as feedback on your writing. Do not write "the reviewer misunderstood". Write something like: "We apologize that this was unclear. We have rewritten the paragraph to state that the test set was never used during model selection." Then change the text. A reader who misunderstood once is a preview of the next one.

When the reviewer asks you to cite specific papers

If the papers are relevant and useful, cite them. If they are not, explain politely why they are not directly relevant to your question. If you believe a request is mainly about citation counts, do not say so in the letter to the reviewers. If the concern is serious, raise it with the editor.

A last read before you submit

• Every comment from every reviewer has a response.
• Each response says what changed and where.
• The page, section, and line numbers match the final file.
• Numbers in the letter match the numbers in the paper.
• Any change you made that no reviewer asked for is mentioned to the editor.
• The tone is neutral from the first page to the last.

From the reviewer's side

I have reviewed more than 50 papers, and a revision looks different from that side. When a reviewer opens your resubmission, the first thing they want to know is whether you took the comments seriously. A letter that is organized, specific, and calm makes that easy to see and saves the reviewer time. A letter that argues with every point makes the second read harder, even when the paper has improved.

My free reviewer rebuttal template, with a phrase bank for agreeing, clarifying, and disagreeing, is on the resources page.