How to Write NIH Specific Aims (With Examples)
The Specific Aims page is the most important page of an NIH application. Here's the structure reviewers expect, common mistakes that sink applications, and how to draft it without tripping NIH's AI rules.
Updated
If a reviewer reads only one page of your NIH application carefully, it's the Specific Aims. It sets expectations for everything that follows and often anchors the score. Getting it right is the highest-value hour you'll spend on the whole proposal.
The structure reviewers expect
Specific Aims pages vary, but the most reliable structure has four moves on a single page:
- The hook (significance + gap). Open with the importance of the problem and the specific gap in knowledge or capability that your work addresses. Make the reviewer care in the first few sentences.
- Hypothesis and objective.State your central hypothesis and overall objective clearly, with the rationale and any preliminary data that make it credible. This is the "why you, why now" paragraph.
- The aims (2–3).Each aim is a distinct, testable objective with a one-line approach. Keep aims independent — don't let Aim 2 collapse if Aim 1 fails.
- Payoff. Close with expected outcomes and the impact on the field if you succeed.
An example aim, weak vs. strong
Weak:"Aim 1: Study the role of protein X in disease Y." — vague, not testable, no approach.
Stronger:"Aim 1: Determine whether protein X regulates pathway Z in model Y. Using [method], we will test the hypothesis that X loss reduces Z activity by ≥50%, establishing X as a candidate target." — testable, specific, with a clear approach and expected result.
Common mistakes that sink applications
- Interdependent aims. If one failure topples the rest, reviewers see risk.
- No hypothesis.A list of activities isn't a research plan.
- Significance assumed, not argued.Spell out why it matters; don't assume the reviewer shares your field's priors.
- Overstuffing. Too many aims, or aims too ambitious for the timeline and budget.
- Vague, unfalsifiable claims."Better understand" is not an outcome a reviewer can score.
Staying inside NIH's AI rules
The NIH has signaled it will refuse applications it considers substantially developed by AI. For the Specific Aims page — your core intellectual contribution — that means the ideas and hypotheses must be yours. AI is genuinely useful for the surrounding work: tightening to one page, checking that each aim is testable and independent, flagging vague language, and confirming you've addressed the review criteria. That's assistive, not generative. Read the details in our 2026 funder AI policy guide, and see the broader case in best AI grant writing tools for researchers.
Write Alongside is built for this: you write the aims, and it rides alongside — checking structure, flagging unfalsifiable claims, and keeping you inside the page limit and the rules. Start freeto draft your next Aims page with a reviewer's eye on it.
Frequently asked questions
What is the NIH Specific Aims page?
It's a one-page summary (typically limited to one page) that states the problem, the gap in knowledge, your central hypothesis, and two to three specific aims — the discrete objectives your project will accomplish. Many reviewers form their initial impression from this page alone, so it disproportionately shapes your score.
How many Specific Aims should I have?
Usually two to three. Each aim should be a distinct, testable objective that stands on its own — not dependent on the success of another aim. Too many aims signals an unfocused project; one aim can look thin.
What's the structure of a strong Specific Aims page?
A common, effective structure: (1) an opening paragraph that hooks with significance and names the gap, (2) a paragraph stating your central hypothesis and overall objective plus your rationale and preliminary basis, (3) the aims themselves with brief approaches, and (4) a closing paragraph on expected outcomes and impact.
Can I use AI to write my NIH Specific Aims?
Use AI to refine and pressure-test your own writing — checking structure, flagging vague claims, tightening to one page — not to generate the science. The NIH has signaled it will refuse applications substantially developed by AI, so keep yourself the author and keep an authorship trail. See our 2026 funder AI policy guide.