NSF Project Pitch Character Limits & Structure

The NSF Project Pitch is a tightly structured document. Each of its four main sections has a fixed character limit — roughly 500 words. There are no images, no tables, no citations, and no formatting tricks. Just text.

That tightness is the whole point. The Pitch is designed to make Program Directors decide quickly. Founders who fight the format lose. Founders who design for it win.

Below: the actual structural rules, what fits in each section, and the editing strategy that makes the limits stop feeling cramped.

Always start with the current portal limits

NSF occasionally updates the Project Pitch field limits. The numbers below are accurate as of the latest published solicitation, but the binding source is always NSF’s SBIR/STTR portal at submission time. Treat this guide as orientation, not law.

Section-by-section structure

1. Technology Innovation (~500 words)

The Pitch reviewer is asking: “What is the new, novel, or unique technical thing here?”

What fits in 500 words:

• 1–2 sentence opener stating the technical innovation in plain language. No company backstory.
• 2–3 paragraphs describing what is technically new — the algorithm, the material, the architecture, the measurable property — versus what is currently state-of-the-art.
• 1 paragraph on why this is genuinely difficult to do (the “risk” that justifies SBIR funding).

What does not fit:

• The company’s founding story.
• A competitive landscape table.
• The product’s UX or branding.
• Commentary on the addressable market (that’s section 3).

2. Technical Objectives and Challenges (~500 words)

The reviewer is asking: “What specifically will you try to accomplish in Phase I, and why is it actually research?”

What fits in 500 words:

• 3–5 specific, measurable technical objectives for the Phase I 6–12 month period.
• For each objective: a one-sentence description, a measurable success criterion, and a one-sentence acknowledgment of why it might fail.
• 1–2 sentences on the technical risks that the work would resolve.

The phrase that gets Pitches rejected at this stage:

“We will build the prototype and validate it with users.”

That’s engineering, not research. NSF wants measurable technical questions whose answers aren’t known. “Achieve X precision under Y conditions while maintaining Z throughput” is a research objective. “Build the MVP” is not.

3. Market Opportunity (~500 words)

The reviewer is asking: “If this works technically, will anyone pay for it?”

What fits in 500 words:

• 1 paragraph on who buys this and why — specific customer profile, specific need, specific willingness to pay.
• 1 paragraph on the addressable market with credible numbers, ideally bottom-up.
• 1 paragraph on the competitive landscape — what people use today and why your approach changes the cost or capability frontier.
• 1–2 sentences on initial customer discovery evidence, if you have any.

What kills this section:

• Quoting “the global X market is $Y billion” with no segmentation.
• Listing every competitor without a sharp positioning argument.
• Conflating “we have a beta user” with “we have a paying customer.”

4. Company and Team (~500 words)

The reviewer is asking: “Are these the right people to do this work, in the right structure, in the US?”

What fits in 500 words:

• 1–2 sentences on the company — incorporation form, location, current headcount.
• 1–2 sentences on US ownership, confirming SBIR eligibility.
• 1 paragraph on the PI — their credentials directly tied to the Phase I work, their primary employment commitment to the company.
• Brief paragraphs on each other key team member with similarly tight credentialing.
• 1–2 sentences on relevant facilities, IP position, or prior funding if it strengthens credibility.

Confirm the eligibility specifics in NSF SBIR eligibility explained before writing this section.

Want this done for you? See our $349 NSF Project Pitch service →

Why 500 words feels impossibly short

The first time most founders try to write Section 1 (Technology Innovation), they produce 1,500 words. They’re trying to do four things at once:

1. Explain the technology.
2. Justify why it’s new.
3. Compare it to what exists.
4. Convince the reader it’s commercially valuable.

Three of those four are the wrong job for Section 1. The market argument belongs in Section 3. The competitive comparison belongs in Section 3. The novelty argument is the only one that needs the words. Once you remove the work that other sections are supposed to do, 500 words becomes generous.

The editing strategy that fits

The strongest Project Pitches are usually written in this order:

1. Draft each section to 1.5x its limit without worrying about length.
2. Cut anything that belongs in a different section. This usually removes 20–30% of each section’s content.
3. Cut anything that’s aspiration rather than evidence. “We will become the leading platform” goes. “We have signed three pilot agreements” stays.
4. Cut adjectives that don’t change the meaning. “Revolutionary,” “cutting-edge,” “world-class,” “disruptive” — always cut.
5. Read it as if you’re a Program Director with 4 minutes. The first sentence of each section should answer the reviewer’s question.

By the time you’re done, each section should be at or just under the character limit, with no filler.

Plain text only — work with it, not against it

The NSF Project Pitch portal does not accept images, tables, or rich formatting. That has consequences:

• No equations. Express mathematical relationships in plain English. “Throughput scales as the square of the number of channels” works. LaTeX does not.
• No diagrams. Replace with structural language: “The system has three subsystems: A, which handles X; B, which handles Y; and C, which handles Z.”
• No headers within a section. Use paragraph topic sentences instead.
• Bullets work, but sparingly. Plain-text bullets (using -) are fine. Prefer paragraphs for substantive arguments and bullets for genuine lists.

Common character-limit mistakes

• Pasting from Word with smart quotes that bloat the count. Some characters take more space than expected. Always run a final paste into the portal and verify the count.
• Long PI biosketch in Section 4. The Pitch is not a CV. One paragraph per key person is sufficient.
• Repeating the technical innovation in Section 2. You introduced it in Section 1. In Section 2, you describe what you’ll do with it.
• Too much commercialization in Section 3. Two paragraphs of TAM/SAM/SOM are enough. The full commercialization plan goes in the Phase I proposal.

A note on consistency between sections

Reviewers read all four sections together. If your Technology Innovation describes a hardware platform but your Technical Objectives are entirely software, that’s a credibility problem. If your Market Opportunity is enterprise sales but your Company section emphasizes consumer expertise, the same. The four sections need to read as one coherent argument.

The cheapest way to check this is to read the four sections aloud, in order, and ask whether they describe the same project to the same reviewer.

Bottom line

The character limits aren’t arbitrary. They force the kind of compression that NSF reviewers value. Sections that fight the limit — trying to do other sections’ jobs, or padding with marketing language — read worse than sections that own the limit and write tight.

If you want sections written by people who do this every week and who know exactly what fits, see our done-for-you NSF SBIR Project Pitch service.