PNAS Response to Reviewers: How to Write a Rebuttal That Wins (2026)

PNAS is not a professional-editor triage journal like Nature, and it is not a biostatistician-gated journal like JAMA. Its defining feature for response-to-reviewers work is the editor-led, returned-to-reviewers model, and that editorial culture shapes what really happens to your rebuttal after you click submit. A member of the Editorial Board owns your paper end to end, chooses the reviewers, and sends your revised manuscript and rebuttal back to those reviewers for a verdict.

Reviews are single-anonymous (single-blind): you do not learn who the reviewers are, but they know who you are and they re-read your paper against your reply.

Two structural facts follow from this. First, the rebuttal is read twice, once by the editor for completeness and once by each reviewer for whether their specific point was met. Write each response to the reviewer who raised it. Second, PNAS sits on a multidisciplinary, broad-interest bar: the 120-word Significance Statement is the triage tool the editor uses to decide whether a non-specialist would care.

When a reviewer says the broad-interest case is not made, that is a substance objection about the paper's reason to be in PNAS, not a cosmetic note. Reworking the Significance Statement and introduction is usually the highest-leverage move in the whole revision.

PNAS also runs a Consultative Review option for dual-classification work that spans two or more major NAS classifications (for example Physical Sciences and Biological Sciences). Papers that draw on adjacent disciplines may be discussed among Editorial Board members in those disciplines, which adds a few days but means your rebuttal can face cross-disciplinary scrutiny. If your paper went through that route, write the response so a reader from the neighboring field can follow why each change matters.

For every comment you mark as resolved, point the reviewer to the exact place in the revised manuscript: page number and line number, plus the figure, table, or reference number when relevant. Write "page 9, lines 6 to 19" or "new Figure 3," not "we have updated the Results." This is not a formatting nicety.

PNAS returns your reply to the reviewers, who re-read the manuscript against it; a response they cannot locate reads as no change and tends to draw the same comment again on the next round, which you may not get. Number your responses to match the reviewers' numbering, and include a tracked-changes file so the reviewer can see exactly what moved.

Make the reviewer's text and your response visually distinct so the editor and reviewers can scan the document fast. Put each reviewer comment in italics or a shaded text box, and keep your response in plain upright text below it. A common, clean convention is bold for the reviewer's comment label, italic for the quoted comment, and regular weight for the author response.

Do not run the comment and your reply together in one paragraph; the reviewer should be able to find their own words at a glance and read straight down to your action.

A revise decision is not an acceptance, and at PNAS the gap between the two is real. Reviewers re-read the revised paper and can recommend rejection if the response did not satisfy them. The honest failure modes:

• Rejection on revision. Multiple revision rounds are rarely permitted at PNAS, and acceptance after revision is not guaranteed. A partial first response, where you defer the hard experiment or only half-answer the significance objection, is the most common way a revise turns into a rejection. Treat the first response as your one full attempt.

• The novelty/general-interest appeal trap. If the paper is rejected, you may appeal in writing within three months.

But PNAS states that appeals on the basis of novelty or general interest are unlikely to be granted, and if an appeal is rejected, the paper may not be resubmitted.

So reframing a substance rejection as a fit argument in your rebuttal usually fails twice: once with the reviewer and again on appeal.

• The two-month clock. A revised paper submitted more than two months after the revision decision may be treated as a new submission, which can mean new reviewers and a fresh round of broad-interest scrutiny. Most authors who miss the window did so because the response stalled on one disputed comment.

Roughly half of PNAS submissions are declined at initial evaluation before peer review, so reaching a revise decision already puts you in a strong minority. Do not waste it with a thin response.

In our pre-submission review work with PNAS submissions, the responses that turn a revise into an acceptance share a shape, and the ones that stall share the opposite shape. In our review of PNAS-targeted resubmissions, three named rejection patterns generate the most consistent second-round trouble, and we observe each one often enough that they are testable against your own response document before you upload it. PNAS editors consistently flag the same three gaps; the editorial triage pattern below is what our review data surfaces most.

The unverifiable acknowledgment. This is the most common pattern across our PNAS reviews. The author writes "we have addressed this concern" or "we have revised the manuscript accordingly" with no page or line number. Because PNAS returns the rebuttal to the original reviewers, who re-read the paper against it, a response they cannot locate reads as no change. We flag every resolved comment that lacks a page/line reference, a figure number, or a reference number. The fix is mechanical: name the exact location for every action verb.

The untouched Significance Statement after a broad-interest objection. When a PNAS reviewer says the work matters only to a narrow specialist audience, the author too often defends the existing abstract instead of rewriting the 120-word Significance Statement and the cross-field framing in the introduction. At PNAS the broad-interest case is the reason the paper belongs in the journal, so leaving it untouched answers the wrong question.

In our reviews, the highest-leverage single edit on a PNAS revision is usually a rebuilt Significance Statement that states what changed and why nearby fields should care, paired with a matching final paragraph in the introduction.

The half-answered control or statistical analysis. The third pattern is deferring the experiment or analysis the reviewer actually asked for. An author will add a paragraph of discussion, propose a control as future work, or report a p-value without the effect size and sample-size rationale the reviewer requested, then mark the comment resolved.

Because multiple revision rounds are rarely permitted at PNAS, a deferred control or an underspecified statistical analysis is one of the most reliable ways a revise becomes a rejection. We flag any response where the action language ("we discuss," "we note") does not match the reviewer's request ("add the control," "justify the sample size").

The fix is to run the analysis, add the new figure or methods text, and cite its location, or to explain with evidence why it is genuinely out of scope and propose a concrete alternative.

A fourth, quieter pattern shows up on cross-disciplinary papers that went through PNAS Consultative Review: the response answers the home-field reviewer well but leaves the adjacent-field reviewer's comment in jargon they cannot follow. When two or more NAS disciplines are in play, write each response so a reader from the neighboring classification can verify the change. Check whether your PNAS rebuttal answers every comment with a verifiable action.

PNAS author guidelines tell you the mechanics: submit the revision within two months, include a tracked-changes file, respond to every comment. They do not tell you what a PNAS reviewer actually does with your reply, or which of your comments will be re-read most skeptically. Pre-submission reviews of PNAS-targeted resubmissions surface the specific failure patterns, the broad-interest substance trap, and the appeal trap that the official guidance leaves implicit.

The PLOS Ten Simple Rules canon below covers universal rebuttal craft; the sections above map that craft onto the PNAS editor-led, returned-to-reviewers model and its broad-interest bar.

• The response addresses every reviewer comment, point by point, with the reviewer's text quoted.
• Each resolved comment names the exact page and line (and figure, table, or reference number where relevant).
• The Significance Statement was reworked if any reviewer questioned the broad-interest case.
• Every requested control, experiment, or statistical analysis was actually performed, not deferred to discussion.
• A clean tracked-changes file accompanies the reply, and you are inside the two-month window.