A major revision is an editorial decision that says: this paper interests us, but it needs substantial changes before we can decide. It is not an acceptance and it is not a rejection. The revised manuscript usually goes back to the same reviewers, who check every change against their own comments.
If it is your first one, the decision letter can read like a polite rejection. The evidence says otherwise: in a published cohort that followed one journal's manuscripts from first submission to final decision, most manuscripts that received a major revision were eventually accepted. This guide explains what the decision means, how it differs from a minor revision and from a reject-and-resubmit, what the data say about your chances, how much time you usually have, and how to organise the response so the second round goes your way.
The editor has read the reviewers' reports and concluded that your paper could be publishable, but not in its current form. "Major" refers to the kind of change requested, not to the number of comments: changes that affect the science of the paper, not just its wording.
In health and behavioural research, the requests behind a major revision usually look like this:
This is the practical difference that shapes your whole response. After a major revision, the revised manuscript and your point-by-point reply usually go back to the same reviewers, who remember their objections and will check whether you did what you say you did. After a minor revision, the editor often checks the changes alone. You are writing for specific people who have already read your paper closely.
Submission systems word it in different ways: "major revisions", "major revision required", "revise and resubmit (major)" or, in some journals, "accept subject to major revision". Whatever the wording, nothing is guaranteed until the editor reads the revised version. Read the decision letter for what matters: the deadline, and the comments the editor singles out.
Almost every journal's decisions fall into five categories. Knowing them prevents two opposite mistakes: celebrating a lukewarm decision too early, and burying a recoverable one.
Very few journals publish what happens to their manuscripts after each decision, so be wary of anyone who quotes a universal figure. The best published data come from a cohort study at the American Journal of Roentgenology (AJR): 696 original research manuscripts submitted in the second half of 2012, followed from first submission to final decision (Rosenkrantz & Harisinghani, 2015).
At first submission, 0.3% were accepted outright, 8.5% received a minor revision, 19.7% a major revision, 65.1% were rejected after review and 6.5% were rejected without review. Of the manuscripts that received a major revision at first submission, 84.7% were eventually accepted, after one or more rounds; for minor revisions the figure was 98.3%.
Read those numbers as an order of magnitude, not as a promise. They come from one radiology journal in one year, and your field, your journal and the content of the reviews all matter. What travels well is the logic behind them: an editor who sends a paper back to the same reviewers is investing their time, and editors do not invest in papers they consider lost. The one sure way to turn a major revision into a rejection is not to resubmit.
The other side is real too. If the new analyses contradict your conclusions, or if the reply avoids the central objections, the paper can be rejected in the second round.
Rosenkrantz & Harisinghani (2015), American Journal of Roentgenology, 204(6), 1152–1156
The deadline is in the decision letter, and it varies between journals; one to three months is a common range. In the AJR cohort, authors took a mean of 73.7 days to send back a major revision, compared with 21.1 days for a minor one (Rosenkrantz & Harisinghani, 2015).
If a new analysis gets complicated or a co-author is unavailable, write to the editor before the deadline and propose a specific new date. Extensions are routinely granted when they are asked for in time, and asking is far better than resubmitting late or half-done.
Rosenkrantz & Harisinghani (2015), AJR
The most common mistake is not a bad reply. It is a badly spent deadline: weeks of paralysis at the start and a response letter written in a hurry at the end. This plan spreads the work over four stages.
Read the editor's letter and every comment once, then close the file. The first reading is emotional: reasonable comments feel like attacks and trivial ones look insurmountable. Two days later, read again with one question in mind: what exactly is each comment asking for? Pay special attention to the editor's letter. When it highlights two or three specific points, it is telling you what will decide the outcome.
Number every comment from every reviewer and sort them into three groups: agree (I will make the change), partly agree (I will do part of it and justify the rest) and disagree (I will not, and I will explain why with evidence). Start the point-by-point response document now, even if it is empty. That table is your project plan: it shows how much real work there is and which comments need new analyses, which are the ones that set your calendar.
Start with the new or repeated analyses. Their results may force you to rewrite the results and the discussion, and you do not want to write those twice. Then restructure sections, then add literature, and leave wording for last. Fill in the response table as you go, with the page and line numbers where each change appears.
Write the cover letter to the editor and check three things: every comment has a reply, every reply points to a change the reviewer can find, and the manuscript contains no change you have not declared. The structure of the reply itself is covered in the guide to writing a response to reviewers.
How to write a response to reviewers
In the second round, reviewers do not reread your paper from scratch: they read your reply with their own list of comments next to it. That is why mistakes of form weigh almost as much as mistakes of substance.
It happens, and it is not a bad sign in itself. In the AJR cohort, only 40.0% to 55.2% of the major revisions sent back at the first and second resubmission were accepted in that round, and yet 84.7% of the papers that started with a major revision were accepted in the end. A second round usually means the reviewers are not yet convinced on one or two points, not that they have changed their minds about the paper.
The risk in the second round is forgetting. You fix the major revision weeks after reading it, and half of the objections have faded. Before you resubmit, put the previous round's comments next to the new version and check, one by one, that each objection is answered in the text and not only in the letter.
Second round of revisions: what you forgot to fix · Rosenkrantz & Harisinghani (2015), AJR
Yes, in most cases. It means the editor considers the paper potentially publishable and is willing to send it back to the reviewers. In a published cohort from one journal that followed every decision, 84.7% of papers that received a major revision at first submission were eventually accepted. It is still not a guarantee: the outcome depends on your reply.
There is no universal figure, because few journals publish it. In a cohort of 696 manuscripts at the American Journal of Roentgenology, 84.7% of those that received a major revision at first submission were accepted in the end, compared with 98.3% of those that received a minor revision. Treat it as an order of magnitude for one journal, not a rule.
The deadline is in the decision letter, and one to three months is a common range. In the AJR cohort, authors took a mean of 73.7 days to resubmit a major revision. If you need more time, ask the editor before the deadline and propose a specific date.
Yes. It happens mainly when the reply avoids the central comments, when the new analyses contradict the conclusions, or when the second reading uncovers a methodological problem the first round missed. A complete, traceable and honest reply makes it much less likely.
Not yet. Some journals word the decision that way, but the paper is only accepted once the editor, and usually the reviewers, have checked the revised version. Treat it exactly like a major revision.
No. You can disagree with any comment if you argue it with data, references or methodological reasons, and the editor acts as referee. What you cannot do is ignore it: every point needs an explicit reply, even if the reply explains politely why you will not make the change.