Building Your EB1A Evidence Package

Why Strategy Beats Volume

If you have already read our 10 EB1A criteria breakdown, you know what USCIS counts as evidence. This chapter is about how to assemble it into a package that gets approved.

Here is the mistake almost every DIY self-petitioner makes: they treat EB1A like a checklist. Find three criteria, dump documents under each, and file. In 2026, that approach is a reliable way to earn a Request for Evidence (RFE) or a denial. Recent adjudication trends show RFE rates on EB1A petitions running roughly 40–50%, and a large share of those RFEs are triggered not by a lack of achievement but by poor evidence positioning — packages that read like a pile of documents instead of a coherent argument.

The petitioners who win think like prosecutors building a case, not like applicants filling out a form. This chapter shows you how.

Policy currency: The framework below reflects the USCIS Policy Manual, Volume 6, Part F, Chapter 2 (Extraordinary Ability), current as of June 2026. See uscis.gov/policy-manual/volume-6-part-f-chapter-2. Note that the two-step framework has been the subject of recent litigation, but it remains the operative adjudication standard USCIS officers apply today.

The Two-Step Kazarian Analysis (and Why "3 Boxes" Isn't Enough)

USCIS evaluates every EB1A petition using a two-step analysis that originates from the 2010 federal appeals decision Kazarian v. USCIS. Understanding both steps is the single most important strategic insight in this chapter.

Step One — The Criteria Count

The officer first checks whether you have submitted qualifying evidence for at least 3 of the 10 regulatory criteria (or a one-time major achievement like a Nobel Prize). This is a largely mechanical, objective count. Does the evidence you submitted for "judging" actually show you judged others' work? Does your "authorship" evidence actually show scholarly articles? If yes to three categories, you clear Step One.

This is the step DIY petitioners obsess over — and it is the easier of the two.

Step Two — The Final Merits Determination

Here is where cases are won and lost. After you clear three criteria, the officer steps back and evaluates the totality of the evidence to answer one question:

Does the entire record demonstrate that you have sustained national or international acclaim and that you are among the small percentage who have risen to the very top of your field?

This is the final merits determination, and it is qualitative, holistic, and discretionary. You can satisfy three criteria on paper and still be denied at Step Two because the record, taken as a whole, does not paint a picture of someone at the top of their field.

The strategic consequences are huge:

• Checking three boxes is necessary but not sufficient. Your package must tell a story, not just pass a count.
• "Sustained" is a real word. A cluster of achievements that all appeared 6–12 months before filing can read as profile-engineering. Acclaim that spans years is far more persuasive.
• The weakest evidence drags down the whole. Padding a criterion with thin documents invites the officer to question your judgment about the entire petition.

Keep both steps in mind as you make every decision below.

Picking Your Strongest 4–5 Criteria

You only need three criteria. You should plan to file four or five.

Why file more than the minimum

Filing a buffer protects you in two ways. First, if the officer disagrees that one criterion is met — a common RFE outcome — you still have three left and your case survives. Second, breadth of recognition (awards and judging and original contributions and publications) reinforces the Step-Two narrative that your acclaim is real and sustained across multiple dimensions.

Do not over-correct into the opposite error. Filing all ten criteria with weak evidence under seven of them is worse than filing five strong ones. Disciplined exclusion of weak evidence is itself a strategy — it signals to the officer that you understand the standard and are not exaggerating.

How to rank your criteria

For each criterion you could plausibly claim, score it on three dimensions:

Rank your candidate criteria and keep the top four or five. A criterion that scores high on all three (e.g., peer-reviewed publications with strong, independent citations) is a load-bearing wall. A criterion that scores low (e.g., an internal "Employee of the Year" award) is a liability — leave it out.

How to Document Each Piece of Evidence

Every exhibit in a strong package has three layers. Submitting only the first layer is the most common reason individually-true claims still fail.

Layer 1 — Primary evidence (the proof)

The official, verifiable artifact: the award certificate, the membership letter, the published article, the patent grant, the journal's reviewer-invitation email. This proves the fact happened.

Layer 2 — Corroboration (the independent witness)

Third-party confirmation that the fact matters: the journal editor confirming you reviewed manuscripts, the association explaining its selective admission criteria, an expert recommendation letter explaining why your contribution was significant. (We cover these in depth in Writing Effective EB1A Recommendation Letters.)

Layer 3 — Context and comparative data (the "so what")

Numbers and benchmarks that show the achievement is extraordinary, not merely competent:

• How many people receive this award each year? Out of how many eligible?
• What is the acceptance rate of this conference or the impact factor of this journal?
• How do your citation counts compare to the median in your subfield?
• What is the typical salary for your role, and where do you fall?

USCIS officers are generalists, not experts in your field. Layer 3 is what translates your accomplishment into a language the officer can evaluate. A document with all three layers is nearly RFE-proof; a document with only Layer 1 is an invitation to question.

Building the Citation and Impact Record

For researchers, engineers, and academics, the citation and impact record is often the backbone of the petition — and the most commonly mishandled.

• Quantity has lost its persuasive power. A long publication list with low citations now raises concerns rather than easing them. A smaller set of genuinely influential papers is stronger. Curate.
• Benchmark your citations. Pull your Google Scholar / Web of Science / Scopus profile, then compare against field norms. "300 citations" means nothing alone; "300 citations, roughly 5× the median for authors in this subfield" is a comparative argument.
• Show downstream adoption, not just counts. Quote later papers that build on your method. Show that a tool you released is used by named labs or companies. Cite a standard, product, or clinical practice that incorporates your work. This is how you prove "major significance" under Criterion 5.
• Document media and non-academic impact. Press coverage about you or your work (not bylined by you), inclusion in industry reports, GitHub stars, download metrics, or product-usage numbers all build the picture.

Comparative Evidence: Proving You're "At the Top"

Step Two literally asks whether you are in the small percentage at the very top of your field. You cannot prove "top" without a comparison. Build comparative evidence deliberately:

• Define your field honestly but precisely. "Machine learning" is too broad to be at the top of; "reinforcement learning for robotics manipulation" is a field you can credibly lead. Define it in your petition letter and then prove dominance within it.
• Use percentiles and ratios, not raw numbers, wherever possible.
• Let independent experts make the comparison for you. A recommendation letter that says "Dr. X is among the three or four people worldwide working at this level on this problem" is comparative evidence an officer can weigh.

A note on comparable evidence (the regulatory term): if a standard criterion does not readily apply to your occupation, the regulations allow you to submit "comparable evidence" of equivalent stature. This is most useful for non-traditional fields (e.g., an entrepreneur, a chef, an open-source maintainer). Use it sparingly and always explain why the standard criterion doesn't fit your occupation before offering the comparable alternative.

Field-Specific Playbooks

The same framework looks different depending on your field. Concrete patterns that work:

Tech / Software Engineering

• Original contributions: patents with evidence of use, widely-adopted open-source projects (download/star metrics, dependent projects), system designs deployed at scale.
• Judging: reviewing for top-tier conferences (NeurIPS, SIGGRAPH, OSDI) or serving on standards committees.
• High salary: total-comp data benchmarked against levels.fyi or a salary survey for your level and metro.
• Leading/critical role: be ready to prove the company itself is distinguished — title alone is no longer persuasive.

Academia / Research

• Authorship + citations as the backbone, benchmarked against subfield norms.
• Peer review for journals and grant panels, confirmed by editor letters.
• Original contributions evidenced by downstream adoption and expert letters.
• Avoid leaning only on letters from your own co-authors and department — officers discount non-independent letters.

Business / Entrepreneurship

• Original contributions: a product, methodology, or company that changed an industry, with revenue, market-share, or adoption data.
• Published material about you: independent press profiles (not press releases you commissioned).
• High remuneration or equity outcomes benchmarked against peers.
• Comparable evidence often carries weight here — frame metrics like ARR, fundraising, or acquisition value with industry context.

Arts / Performing Arts

• Display/exhibitions at recognized venues, with catalogs and critical reviews.
• Awards with selectivity data.
• Commercial success: sales, streaming numbers, box office, ticket data with venue prestige.
• Published material: independent reviews and features in recognized outlets.

The Most Common Evidence Weaknesses That Trigger RFEs

Based on current 2026 adjudication trends, these are the avoidable patterns that draw RFEs:

1. A disconnected pile of documents. Individually fine, but no narrative ties them to "top of the field." Fix: a petition letter that connects every exhibit to the Step-Two story.
2. Quantity over quality — long publication lists with thin citations; many weak awards. Fix: curate to your best, cut the rest.
3. Missing Layer 3 (context). Claims with no comparative benchmark. Fix: add selectivity rates, percentiles, impact factors.
4. Non-independent or generic recommendation letters — all from co-authors, colleagues, or your employer; or templated language. Fix: independent experts who can name your specific impact.
5. "Distinguished organization" not proven. Leaning on a senior title without showing the org's national/international standing.
6. Profile engineering. A cluster of awards, memberships, and media all dated months before filing. Fix: emphasize the sustained span of your record.
7. Poor organization. Excessive narrative, unlabeled exhibits, evidence the officer can't quickly verify. Fix: a clean exhibit list (below).

If you do receive an RFE, do not panic — it is often recoverable through re-positioning rather than new achievements. See Responding to an EB1A RFE.

Organizing Exhibits and the Exhibit List

Make approval easy. An officer who can verify your claims in seconds is an officer inclined to approve.

• Number every exhibit (Exhibit 1, Exhibit 2, …) and reference each one by number in your petition letter.
• Lead with an exhibit list — a table of contents mapping each exhibit to the criterion it supports.
• Group by criterion, then within each group lead with primary evidence, then corroboration, then context.
• Tab and label so the officer never has to hunt.
• Translate all non-English documents with a certifier's statement.

A simple exhibit-list format:

Your Evidence-Package Checklist

Before you file, confirm:

• 4–5 criteria documented (a buffer above the minimum of 3), each genuinely strong.
• No weak/padding criteria that could undermine your credibility.
• Every exhibit has primary evidence, and the important ones have corroboration + context (Layer 3).
• Comparative data (percentiles, selectivity rates, citation benchmarks) for your key achievements.
• Independent expert letters — not just co-authors or your employer.
• A petition letter that explicitly argues the final merits Step-Two story: sustained acclaim + top of the field.
• Your field is defined precisely enough to credibly be at the "top" of it.
• Evidence shows a sustained record over time, not a last-minute cluster.
• A numbered exhibit list / table of contents, with all foreign documents translated.
• You have re-read the package as the officer will — can a non-expert verify each claim in seconds?

Get these right and you are not just clearing three boxes — you are handing the officer a complete, coherent case for why you belong among the very few at the top of your field.

Building your package and not sure which criteria are truly your strongest? Start with a free evaluation — a second set of eyes on your evidence before you commit can be the difference between a clean approval and an RFE.