NIH R01 Approach Section: What Reviewers Actually Score

Why Approach Carries More Weight Than You Expect

NIH has published data showing that of the five peer review criteria, Approach correlates most strongly with the overall impact score. This held under the old five-criterion scoring system and it holds under the simplified review framework that took effect in 2026, which reorganized those five criteria into two scored factors plus an Environment sufficiency check. Factor 2, which covers Approach, routinely determines whether a competitive Significance score gets translated into a fundable application or sits at the edge of consideration for another cycle.

The funding climate has made this more consequential. Success rates for most R01 mechanisms have trended down over recent years, and when success rates fall, reviewers become more granular in Approach critiques. It's not that they're looking for reasons to reject your idea. It's that a study section panel needs to rank a set of applications that are all scientifically interesting, and the tool they use to discriminate between them is the quality of the experimental plan. If two applications have similar Significance scores and similar biosketches, Approach is almost always what separates funding from a request to resubmit.

The upside is that Approach is the section you control most. Your Significance score depends partly on your field's strategic position in NIH's current portfolio priorities. Your Innovation score depends on what other people have already published in your area. Approach is a writing and planning problem, which means it responds to deliberate work in a way that the other sections don't.

Structure: How to Organize the Section Across Multiple Aims

The Approach section doesn't have a mandatory format, and that freedom causes more problems than it seems like it should. The structure that works for most R01 applications is: a brief conceptual overview, then one clearly labeled subsection per aim, then a timeline or feasibility statement, then a brief closing note on overall rigor.

The conceptual overview is not a second abstract. It's two to four sentences that tell the reviewer how the aims fit together as a research program — what the logical through-line is and why the experiments are in the order they're in. This paragraph matters because reviewers often move from the Specific Aims page directly to the Approach section. They need a moment to reorient before they encounter detailed methods. If the first thing they hit is an assay description, they're already slightly behind, and that mild disorientation can shade the tone of their notes.

Within each aim's subsection, a structure that holds up is: rationale (one paragraph), experimental design (two to four paragraphs), expected outcomes (one paragraph), and potential pitfalls with alternatives (one paragraph). Resist the temptation to vary this from aim to aim. Reviewers read multiple applications in the same sitting. A consistent format lets them find the expected-outcomes paragraph without hunting. When they have to hunt, they write clarity critiques, and those end up in your summary statement.

Don't open the Approach section with a literature review. If your Significance section did its job, reviewers already understand the landscape. Starting Approach with another three paragraphs of background reads as filler and eats page space you need for experimental detail.

Writing Experimental Design Without Overdoing It

The most common failure in Approach writing isn't being too vague. It's oscillating between vague and excessive within the same section. A paragraph that says "we will use standard immunohistochemistry" is too thin. A paragraph that lists every antibody dilution and antigen retrieval temperature belongs in a methods appendix. Reviewers need something in between, and finding that middle ground is mostly a question of what you're being specific about.

What reviewers actually need in the experimental design is specificity of decision and rationale. Tell them which assay you'll use and why it's appropriate for your specific research question. Tell them your key timepoints or conditions and why you chose those. Describe how you'll handle the most likely technical failure mode for that assay. That's enough. You don't need to teach reviewers how the assay works; you need to convince them you know how to use it for this problem.

Sample size and statistical power belong in the experimental design subsection, not in a generic statistics paragraph tacked onto the end of the Approach section. Reviewers check for power calculations, and when they don't find them near the relevant experiment, they flag it explicitly in the criterion notes. A single sentence per aim is usually enough: name the comparison you're powering, the expected effect size and its basis, and the resulting sample size. If your field is early enough that you don't have a good effect size estimate, say so and propose how you'll get one in the first aim. That's better than silence.

You don't need to justify every standard method. If you're doing RNA sequencing to measure transcriptomic changes, don't explain what RNA sequencing is. Use that space to explain why you chose bulk over single-cell for this aim, what your depth target is, and what your differential expression threshold will be. Justify decisions, not techniques.

Rigor, Reproducibility, and Sex as a Biological Variable

NIH has required attention to rigor and reproducibility since the 2016 policy update, and reviewers still flag failures here in a meaningful fraction of summary statements. Under the simplified review framework, the Approach factor score explicitly encompasses rigor, so there's no ambiguity about where this lands.

Each of these deserves a clear statement somewhere in the Approach section, but don't bolt on a dedicated "Rigor and Reproducibility" paragraph at the end. That signals to reviewers that you treated it as a compliance checkbox rather than a design principle. Instead, address the scientific premise in your opening conceptual overview, integrate blinding and randomization into each aim's experimental design paragraph, put a one-sentence authentication statement where you first introduce a key reagent, and handle sex as a biological variable either in the study design or in a brief note that's specific to your experimental system. Silence on any of these four components is almost always noted.

Alternatives and Contingency Plans: How Much Is Enough

Every Approach section needs alternatives. Most PIs either skip them entirely or write a single paragraph at the end that reads as a generic disclaimer. Neither works.

The version that lands with reviewers is one specific alternative per aim, placed at the end of that aim's subsection, tied directly to an anticipated failure mode. The structure is: if a specific thing doesn't work, you'll use a specific alternative approach, for a one-sentence reason. That's the whole formula.

What makes an alternative credible is specificity. "If this aim encounters difficulties, we will adjust our experimental approach as needed" tells reviewers nothing useful. "If lentiviral transduction efficiency falls below 50% in primary neurons, we'll switch to the lipofectamine protocol we validated in our 2024 paper" tells reviewers that you've actually thought about failure modes and have the resources ready. That's the distinction reviewers respond to.

Avoid writing alternatives for disasters with a low probability of occurring. Focus on the one or two failure modes per aim that are genuinely most likely given your current preliminary data. Reviewers recognize when a contingency section is being padded, and excessive alternatives can actually lower your Approach score by suggesting you expect the original plan to fail. One credible alternative per aim is better than four vague ones stacked into a block paragraph at the section's end.

Figures, Tables, and the Page Limit Math

The Research Strategy section has a 12-page limit for most standard R01 applications, and the Approach section typically consumes eight to nine of those pages once Significance and Innovation are written. That means you have genuine constraints on how many figures you can include before they start crowding out the prose needed to explain your experimental plan.

A useful rule of thumb: one figure per aim, plus a Gantt chart if you have room. Figures should earn their space by conveying something that would take more than half a page to explain in prose — a schematic of your experimental workflow, a key piece of integrated preliminary data, or a decision tree for how you'll interpret ambiguous outcomes from a critical experiment. A figure that just illustrates what the prose already says clearly is wasted space.

Tables work well for animal group assignments, sample comparisons across conditions, or multi-assay timelines where the tabular structure genuinely helps the reader scan. They're not a good fit for lists of methods or for anything that's more readable as running text.

A timeline Gantt chart at the end of the Approach section is optional but often worth including for first R01 applications. Reviewers checking feasibility — can this team actually complete all of this in five years? — can answer that question in ten seconds with a well-designed Gantt chart, versus several minutes of adding up experiment timelines scattered across three subsections. For experienced investigators submitting competitive renewals, the timeline is usually less necessary, because the biosketch already makes the productivity case.

Frequently Asked Questions