NIH Research Strategy Figures and Formatting: Make Your Pages Easy to Score

Why Layout Changes How Reviewers Score

Reviewers won't admit it, but they skim. A typical assigned reviewer handles several applications per cycle, and the Research Strategy is the longest piece of each. Under the simplified review framework, they're rating the importance of the research and the rigor and feasibility of the approach, and both judgments depend on finding evidence quickly. If your key result is buried in a dense paragraph on page 7, it effectively doesn't exist.

I'm not arguing that formatting can rescue weak science. It can't. The argument is narrower: a strong application that's hard to navigate gets scored as if it were slightly weaker, because confusion reads as risk. A reviewer who has to hunt for the sample size, the control group, or the preliminary result supporting Aim 2 starts writing "unclear" in their critique. That word is rarely good for an impact score.

Spending Your Figure Budget Where Reviewers Look

Think of the 12-page limit for R01-sized Research Strategy sections as a budget. Every figure costs space, and every figure should earn that space by doing a job text can't do as well. Three jobs justify a figure: showing the preliminary result that makes your hypothesis credible, showing a conceptual model so the panel can hold the whole project in their heads, and showing a study design or timeline that would take a paragraph to describe.

A good rule of thumb is to put your single most persuasive preliminary data figure at the top of the Approach for each aim, not tucked in the middle of a methods paragraph. Reviewers tend to look at pictures first and read the text second, so the figure sets the frame. If you have only room for one conceptual schematic, place it in the Significance or Innovation section, where it can orient the panel before the technical detail arrives.

Resist the urge to cram six panels into a half-page figure. Text inside a panel that's unreadable at 100 percent zoom is worse than no figure. If a reviewer has to zoom in to read axis labels, you've already lost some goodwill. Aim for labels that stay legible when the PDF is viewed at normal screen size, and cut panels that don't serve the argument for that aim. Supplementary panels belong in the text as one-line summaries, not as thumbnails nobody can read.

Legends That Carry the Argument

Legends are the most underused real estate in a grant. Many applicants write them like journal captions: a title, then a list of methods. A grant legend has a different job. The first sentence should state the conclusion the figure supports ("Knockdown of X reduces tumor burden by roughly half in the orthotopic model"), and only then should it give the n, the statistics, and the essentials of the design. A reviewer who reads nothing but the figure and its first legend sentence should still get your point.

Put the sample size and the test in the legend, every time. Rigor is part of how reviewers judge feasibility, and a figure with no n invites doubt. If the preliminary data come from a pilot with small numbers, say so plainly and explain in the text how the proposed aim scales it up. Reviewers forgive an honest small pilot much more readily than an implied big one that turns out not to be.

Legends can be set in a slightly smaller font, as long as you stay within NIH's font and size requirements. Don't push this. Check the current NIH formatting guidance for the minimum size allowed in figures and tables, and keep a margin of safety, because the submission system can flag problems you didn't see in your word processor.

Headings, Bold Text, and Signposting

Use informative headings, not generic ones. "Preliminary Data" tells a reviewer nothing. "Preliminary data: loss of X sensitizes cells to Y in three models" tells them what they're about to learn and lets them skim past if they already believe it. The same applies at the aim level. A sentence-style heading that states the hypothesis, followed by the rationale, the approach, expected outcomes, and potential pitfalls, gives each aim a predictable skeleton.

Predictability is the point. If Aim 1, Aim 2, and Aim 3 each follow the same order, reviewers learn the pattern in a page and then navigate by habit. Bold the one or two sentences per section you most want remembered, such as the central hypothesis or the key decision rule that determines whether you move on. Bold used on every third sentence is just noise, though. If everything is emphasized, nothing is.

It's also worth ending each aim with a brief statement of what a positive and a negative result would mean for the project. That's the kind of sentence a reviewer can lift straight into their critique, and it quietly answers the "what if it doesn't work?" question before it's asked.

White Space Is Not Wasted Space

Applicants who are short on room often fill every line, shrink every margin they dare, and produce walls of text. Reviewers read those walls with visible fatigue. Short paragraphs, a line of space between them, and a figure every page or two make the same content feel lighter, even when it isn't shorter.

A practical check: print the Research Strategy or view the PDF zoomed out so you can see a full page at once. Squint. Can you tell where the figures are, where the headings are, and where each aim starts? If the pages all look like gray rectangles, a reviewer will experience them that way too. I usually recommend cutting about 10 percent of the words to win back white space, since the weakest 10 percent rarely earns a score anyway.

Tables deserve a mention here. A table is a good way to show aims against methods, or timeline against milestones, because it compresses information that would otherwise sprawl. But a table is also a magnet for tiny type. Keep it small enough to fit and big enough to read, and don't add one just to look organized.

Formatting Mistakes That Cost You Credibility

One more caution: the rules around attachments, links, and appendix material have changed in recent years, and I'm not going to claim I've memorized the current details. Verify against the current NIH application instructions for your specific form and funding opportunity before you finalize, because an individual announcement can add its own limits.

A 30-Minute Figure Audit

Once the draft is stable, run this audit before anyone outside your lab sees it. It takes about half an hour for a typical R01-sized Research Strategy.

• One-sentence test: for each figure, write the point it makes in one sentence. If you can't, cut or redesign it.
• Legibility test: view the PDF at normal screen size and check every label without zooming.
• Rigor test: confirm every data figure states n, the statistical test, and what the error bars represent.
• Placement test: confirm each aim opens with or soon reaches its strongest supporting figure.
• Skim test: hand the PDF to a colleague outside your subfield for two minutes and ask them to describe the project back to you.

If you're also running an internal critique session, fold this audit into it so reviewers comment on readability as well as science. The two problems are connected more often than people expect.

Frequently Asked Questions