Hundreds of Proposals Written
Federal, State & Foundation Grants
SBIR/STTR, Grant Strategy

AFWERX SBIR Eligibility Requirements: The 10-Gate Compliance Pre-Screen

Cada has written 100+ grant proposals across 30+ agencies, including AFWERX, DLA, and DoD.

Most defense SBIR advice is about how to win. This is about the ten ways you get eliminated before a reviewer opens your proposal.

None of them touch your technology. Your ownership structure, your PI's other job, your subcontract split, and the currency of five federal registrations decide whether you get evaluated at all.

This is the Phase 0 pre-screen Cada runs before agreeing to draft for a defense client. Run it on yourself first. Twenty minutes now is cheaper than the 40 to 80 hours of writing you spend on a proposal you turn out to be unable to submit.

What are the AFWERX SBIR eligibility requirements?

An AFWERX SBIR applicant must be a US for-profit small business with no more than 500 employees, more than 50% owned by US citizens or permanent residents, and led by a PI employed more than 50% by your company. All R&D happens in the US. Five registrations must be current, and your ITAR, FOCI, and Section 889 posture clean.

How to run this DoD SBIR compliance checklist before you write anything

Mark each gate one of three ways:

  • CLEAR -- you meet it today, with documentation.
  • NEEDS_ACTION -- fixable, but not fixed. Has a lead time.
  • CRITICAL -- you do not meet it and cannot without restructuring the company or the team.

Then apply two rules.

Any single CRITICAL stops you. Compliance is not a majority vote. Nine clean gates do not offset one failure.

Two or more NEEDS_ACTION means fix your posture before drafting, not alongside it. An AFWERX Open Topic Phase I is $75,000 over 90 days. Writing 40 to 80 hours while two registrations are pending is how founders end up watching a deadline pass with a finished proposal they cannot submit.

The 10 AFWERX SBIR eligibility requirements, gate by gate

Gate 1: Company size and structure

For-profit, organized in the United States, no more than 500 employees. The count is an average over the preceding 12 months, and it includes affiliates.

The affiliate rollup is where this bites. If a parent or a common owner controls other entities, those headcounts roll up into yours. The applicants who fail here are usually subsidiaries who never thought to count the parent, and they find out after the size protest, not before.

Gate 2: US ownership, and the venture capital trap

More than 50% must be owned by US citizens or lawfully admitted permanent residents. Exactly 50% is not enough. It has to be more than half.

Here is the part that catches funded startups. Under SBA rules, agencies may opt in to funding firms majority-owned by multiple VC, hedge, or private equity funds. NIH opted in. DoD did not. A single fund holding more than 50% on its own disqualifies you at every agency, opt-in or not.

So a company fully eligible for an NIH SBIR can be flatly ineligible for an AFWERX one, with no change to the technology. If your cap table crossed 50% institutional ownership last round, this is CRITICAL, and the fix is months of cap table work, not a rewrite.

Gate 3: Principal investigator primary employment

Your PI must be employed more than 50% of their total working time by your company. Not 50% of the project budget. More than half of everything they do for a living.

A full-time professor who spends 20 hours a week on your company fails. A PI splitting 40% and 40% across two startups fails. DoD applies this at award and throughout performance, so a PI who returns to a faculty appointment mid-project is a problem you inherit later.

Gate 4: The registration stack and the CAGE code SBIR requirement

Five registrations, in dependency order: SAM.gov, which issues your UEI and triggers your CAGE code, then SBIR.gov, then DSIP.

The CAGE code SBIR requirement has the ugliest lead time. The Defense Logistics Agency assigns it during SAM.gov registration, and DLA publishes a 10-business-day standard that it has been running past. A new SAM.gov registration takes three to four weeks end to end.

Renewals matter too. SAM.gov expires every 12 months, and an expired registration reads the same as no registration. Check your expiration date now, not the week the solicitation closes.

Gate 5: Place of performance

All research and development must be performed in the United States. Not most. All.

This is a quiet CRITICAL for teams with an offshore engineering arm. If real technical work happens abroad, restructure the plan before writing, because the budget and the work plan both have to show US performance.

Gate 6: ITAR and EAR posture

Defense SBIR ITAR compliance is where founders assume they are fine because nobody told them otherwise.

ITAR (22 CFR 120-130) controls defense articles, technical data, and services on the US Munitions List. EAR covers dual-use items. If your technology is controlled, foreign national participation is restricted and must be disclosed.

Founders sometimes reach for the fundamental research exclusion. Be careful.

That exclusion is built around research at accredited US universities intended for public release. SBIR work at a company, aimed at a defense end-user and not meant for publication, usually does not fit.

This is the one gate you should not self-certify. If there is any chance your technology is controlled, the honest status is NEEDS_ACTION pending export counsel, two to six weeks.

Gate 7: FOCI and foreign disclosure

FOCI SBIR small business review is the gate founders misunderstand most.

FOCI means Foreign Ownership, Control, or Influence: a foreign interest holding power to direct or decide matters affecting your management or operations. DoD requires disclosure of foreign relationships for all owners and covered individuals, meaning senior key personnel.

AFWERX evaluates four areas: cybersecurity practices, patents, employee affiliations, and foreign ownership. The countries of concern are China, Iran, North Korea, and Russia.

Here is what most founders get backwards. Disclosing a foreign relationship does not disqualify you. Failing to disclose one does. AFWERX says plainly that disclosure is not itself disqualifying, and equally plainly that an association discovered later, including after award, can mean disqualification, termination, and referral.

One foreign angel investor, disclosed, is usually NEEDS_ACTION. The same investor, undisclosed, is existential.

Gate 8: NDAA Section 889 telecommunications

You must certify that you do not use covered telecom equipment from Huawei, ZTE, Hytera, Hikvision, or Dahua, including their subsidiaries and affiliates, as a substantial or essential component of any system.

Check your hardware, not your intent. Security cameras and network gear are the usual culprits, and Hikvision and Dahua cameras are frequently white-labeled under other brand names. Verify the manufacturer, not the sticker.

This gate is cheap to clear and expensive to get wrong. Replacing a camera costs a few hundred dollars. Certifying falsely is a false statement on a federal award.

Gate 9: The AFWERX subcontract percentage limit

For SBIR Phase I, your company must perform at least two-thirds of the work by cost. Subcontractors and consultants together are capped at one-third.

Two things founders get wrong. First, consultants count: a 1099 advisor and a subcontracted lab land in the same bucket.

Second, and this one is genuinely obscure, DoD subtracts profit and fee before running the percentage. Your denominator is not the grand total on your cover sheet. The calculation is:

(Subcontracts + Consultants) / Total Cost Before Fee <= 0.33

A budget that passes against the grand total can fail against cost before fee. Run it both ways before you commit to a work plan, because moving 5% of scope in-house after the budget is built means rewriting the technical approach too.

STTR runs on different math: your company at least 40%, the research institution at least 30%.

Gate 10: Budget conformity

AFWERX Open Topic Phase I is capped at $75,000 with a 90-day period of performance, STTR at $110,000. Over ceiling is not a negotiation opener. It is screened administratively, the same as a missing registration.

On profit and fee: 5% to 7% is the normal range, and rates above 7% invite negotiation. To be clear, that is a practical ceiling drawn from what gets accepted, not a published regulatory cap. Treat 7% as the point where you should expect questions, not a rule you can be disqualified under.

What to do if you fail two or more gates

Fix in order of lead time, not severity. The slowest thing you start today finishes first.

  1. Registrations (weeks). Start SAM.gov immediately. Nothing downstream moves without it.
  2. Export counsel (two to six weeks). Start in parallel if Gate 6 is open.
  3. Structural fixes (months). Cap table, PI employment, work-share.

Consider a fictional six-person autonomy startup. The PI is still 60% employed at a university, and 45% of the budget goes to that same university lab. Two gates fail.

Nothing in the technical plan is wrong, and no amount of rewriting fixes it, because the problem is the org chart. Restructure this cycle, submit to the next. That is a worse answer than "you're ready," and it is a far better answer than finding out in month four.

Get the fillable version

The rule set above is complete. Nothing is held back.

What is gated is the working copy: a fillable 10-gate worksheet with status boxes, the work-share and budget arithmetic worked line by line, a fix list sorted by lead time, and the STTR variants. Fill it in and you will know in 20 minutes whether to write.

Two or more CRITICAL flags is not a proposal problem, and a grant writer will not solve it. It is structural: your cap table, your team, your work plan.

If that is where you land, book a free 20-minute compliance-posture call rather than asking for a proposal quote. Bring your cap table summary, your PI's employment split, and a draft budget. You leave knowing which gates are actually blocking you and what order to fix them in. No pitch, no obligation.

Sources

Eligibility rules, caps, and disclosure requirements are set per solicitation and change between cycles. DoD volume structures and disclosure forms in particular change between releases, so verify against the solicitation you are applying under before you plan a restructure. Nothing here is legal advice, and Gate 6 in particular requires export counsel rather than a checkbox. All company examples are fictional and used for illustration only.

Frequently Asked Questions

Compliance and registration failures are screened administratively, so a proposal that fails them may never reach technical evaluation. Some checks continue after submission and after award, including foreign risk analysis. An undisclosed foreign association discovered post-award can result in disqualification, termination of the award, and referral for further review.
You can generally submit with a registration in progress, but no award can be made without an active SAM.gov registration. A new registration takes three to four weeks end to end, because the CAGE code assigned by DLA carries its own 10-business-day standard. A pending registration near a deadline is a real risk.
The Defense Logistics Agency assigns your CAGE code during SAM.gov registration, so you do not request it separately. DLA publishes a 10-business-day standard and has been running past it. Budget three to four weeks for a new SAM.gov registration end to end, and start it before you write anything.
No. Foreign investment triggers disclosure and FOCI review, not automatic disqualification. What disqualifies you is majority foreign ownership or control, or failing to disclose a relationship that is discovered later. Disclose every foreign relationship for all owners and covered individuals, meaning senior key personnel, when you apply.
For SBIR Phase I, subcontractors and consultants combined are capped at one-third of the work by cost, and your company must perform at least two-thirds. DoD calculates this on total cost before profit and fee, not on your grand total. STTR uses a different split: at least 40% company, at least 30% research institution.

Ready to explore your funding options?

We'll map your technology to the most relevant programs and tell you where to start. 15 minutes, no obligation.

Book Strategy Review