NIH Simplified Review Framework: One Year of Reviewer Feedback and What to Change Now

Why the Framework Changed the Review Conversation

NIH's simplified peer review framework took effect for applications due January 25, 2025. By August 2026, at least three full R01 review cycles have completed under the new rules — enough time for patterns to emerge, for reviewers to report honest observations, and for applicants who submitted early to see whether what they assumed about the new structure was correct. For many, it wasn't.

The core reorganization is worth a quick recap. NIH kept the five statutory criteria — Significance, Investigators, Innovation, Approach, and Environment — but changed how reviewers score them. Factor 1 combines Significance and Innovation into a single numerical score. Factor 2 is the Approach score, unchanged in concept but now carrying more relative weight in the two-score world. Factor 3 covers Investigators and Environment but is no longer scored numerically: reviewers rate it as either sufficient or insufficient, not with a 1-through-9 number. Overall Impact remains the primary score that drives discussion and eventual funding decisions.

Two effects applicants didn't fully anticipate: Factor 3 shifted from a place where you could earn points to a place where you can lose the application. And the combined Factor 1 score changed what reviewers actually reward in the first section of your Research Strategy — not by changing the criteria, but by changing how reviewers reason about them together.

What NIH Heard from Reviewers After the First Full Year

In January 2026, the NIH Extramural Nexus published a piece titled "Learn What Reviewers Say About Implementing the Simplified Review Framework." It drew on study section feedback collected across the first several cycles under the new rules. A few findings stand out for applicants.

Reviewers generally found the two-score structure easier to work with at the panel. Fewer numbers to defend meant faster consensus on borderline applications. Discussion time for applications near the funding threshold got shorter — which sounds like a neutral observation but isn't. If your application reaches discussion, your advocate has less time to build the case than they might have had under the old framework. The Specific Aims page matters even more as a result, because it's often what the advocate leans on when the floor discussion is compressed.

Reviewers also reported that the separation between Factor 1 and Factor 2 was useful for identifying a particular class of application: strong scientific problems, well-framed, but with experimental designs that didn't hold up under scrutiny. Previously, a high Significance score would partly offset a weak Approach score in reviewers' mental models, even informally. The simplified structure made that offset less automatic. A strong Factor 1 couldn't rescue a weak Factor 2 the way it might have before.

The third finding surprised program officers. Factor 3 sufficiency determinations were generating more discussion at the panel than expected. Reviewers were uncertain about when to mark something as insufficient, particularly for early-career investigators and multi-institutional teams. Applications that treated Factor 3 as an afterthought were occasionally getting flagged, even when the science was strong. This created a new category of reviewer concern that didn't have a clean analog in the previous framework.

Factor 1: You Still Can't Write Significance and Innovation as Two Separate Things

The most common early mistake was submitting Research Strategies where Significance and Innovation were clearly written as distinct sections that happened to share a score. The prose would move through three or four paragraphs on why the problem matters, hit a subheading, and then pivot to explaining what was new about the proposed approach. Reviewers assigned a single Factor 1 number, but the underlying argument was organized as if they were still scoring two separate criteria.

This matters because a combined Factor 1 rewards a specific kind of scientific argument: one where the importance of the problem and the novelty of your approach to it are genuinely intertwined. The strongest Factor 1 cases explain why the field gap exists specifically because no one has tried what you're proposing. The weakest present a genuinely important problem and then, a few paragraphs later, explain that you're doing something new — with no logical thread connecting the two.

If your current draft has a paragraph that starts with something like "To address this important gap, we propose a novel approach..." the structure probably isn't there yet. The connection needs to be built into how you frame the problem, not added as a transition sentence. Reviewers scoring a single Factor 1 number are asking one question: does the significance of this problem and the novelty of this approach reinforce each other? If they have to mentally connect those arguments themselves, the score suffers.

Factor 2: The Approach Score Now Has More Influence on Overall Impact

Under the old five-criterion framework, an application with a mediocre Approach score but strong Significance and Innovation could still earn a competitive Overall Impact score. Reviewers were averaging across five numbers, and a 2 on Significance could absorb a 4 on Approach in a way that kept the overall impression favorable. The simplified framework's structure doesn't work that way.

With only two numerical criterion scores, the Factor 2 Approach score accounts for a larger share of the variation in Overall Impact. Reviewer feedback from the first year confirmed this wasn't subtle. Applications where the two factor scores diverged significantly — a strong Factor 1 paired with a weak Factor 2 — were consistently not discussed. Applications where Factor 1 and Factor 2 were close in score were pulled into discussion more often, even when neither score was particularly strong in absolute terms.

The practical upshot: your Approach section needs to succeed on its own, not as supporting documentation for a compelling Significance argument. Reviewers need to believe the experimental design is feasible, that the timeline is credible, that the alternatives are genuinely alternative designs rather than defensive hedges, and that the rigor requirements are fully addressed. Preliminary data plays a bigger role in establishing Approach credibility under the simplified framework than it did before, especially for new investigators who don't yet have a track record of delivering on similar aims. The data you include needs to speak directly to the feasibility of what you're proposing, not just to the general area of your work.

Factor 3: What "Sufficient" Actually Requires

NIH designed the Factor 3 sufficiency determination to reduce reviewer burden on the Investigators and Environment dimension. Instead of debating whether a team deserves a 2 or a 3 on a nine-point scale, reviewers now make a binary call. The guidance NIH gives reviewers frames sufficiency as whether the investigators have the expertise, experience, and resources to carry out the proposed research. That sounds like a low bar until you see how it plays out.

An insufficient Factor 3 determination doesn't produce a numerical penalty — it creates a different kind of signal in the reviewer write-up, one that program officers and institute staff read carefully. An application flagged as insufficient on Factor 3 can still receive favorable Overall Impact scores and can be funded, but the funding decision is harder for a program officer to defend when the reviewer record includes an explicit sufficiency concern. It becomes a conversation item at the portfolio level that wouldn't otherwise exist.

The applications drawing Factor 3 flags in early cycles had one thing in common: boilerplate environment descriptions that didn't connect to the specific needs of the proposed aims. A general statement about institutional resources at a research-intensive university tells reviewers nothing about whether you have the equipment, access, and collaborators this particular project requires. One or two specific sentences — naming the core facility, the patient population access, or the collaborator bringing a technique your lab doesn't have — is enough to satisfy most panels. The goal isn't a long section; it's a targeted one.

Five Adjustments Before Your Next Submission

Frequently Asked Questions