NIW Recommendation Letters: Independent vs Dependent + Structure

Recommendation letters are the most labor-intensive component of an NIW petition — and the component most often done wrong. A petition with strong publications and a good proposed-endeavor statement fails when the letters are generic, from the wrong people, or don't map to the three Dhanasar prongs.

Here is what strong letters look like, who should write them, and what gets petitions denied even when the underlying record is solid.

The independent vs dependent distinction

USCIS distinguishes between dependent and independent recommenders. Dependent recommenders have a direct professional relationship with the applicant: PhD supervisors, postdoc mentors, co-authors, colleagues in the same lab or department, professional collaborators.

Independent recommenders know the applicant's work from the literature, from professional reputation, or from industry — without a direct working relationship. They cite the applicant's papers; they have used or built on the applicant's methods; they have read and evaluated the work from the outside.

The practical benchmark: aim for at least 60% of your letters from truly independent experts. In a seven-letter petition, that's four or five independent letters and two or three dependent. Many strong petitions do better — six of seven independent.

This is not a formal USCIS rule with a specific percentage threshold. It is the practice standard that reflects how adjudicators weight the letters.

Mapping letters to the three prongs

Every letter should address prong 2 (well-positioned to advance the endeavor) — it's the core function of a recommendation letter. But a well-designed letter set also covers the other prongs:

Prong 1: Substantial merit and national importance

One or two letters should explicitly address why the applicant's proposed work area has national importance — ideally from recommenders with specific expertise in the policy or applied context of the field (a government scientist, an industry researcher in a critical technology area, a researcher at a national lab). Generic statements ("this field is important") don't advance prong 1; specific, documented connections to national priorities do.

Prong 2: Well-positioned

This is the main job of most letters. Recommenders should describe: what they know about the applicant's work (specific papers, methods, contributions), how they became aware of the work (indicates independence), their assessment of the applicant's place in the field relative to peers, and why they believe the applicant specifically is capable of executing the proposed endeavor.

Vague superlatives ("one of the best in the field") without specificity do not advance prong 2 effectively. What works: specific comparisons ("among the top 5% of researchers I have evaluated in this subfield"), specific citations to the applicant's work, specific descriptions of impact or adoption.

Prong 3: On balance beneficial to waive

Letters from U.S. institutional representatives (university research directors, national lab staff, industry technical leads) who specifically want the applicant to continue working in the U.S. are the strongest prong-3 evidence. These letters should state explicitly: why the U.S. entity benefits from this applicant's continued presence, why the work cannot wait for a PERM labor certification, and what specific programs or initiatives depend on the applicant's skills.

What goes in the letter: structural outline

A strong NIW recommendation letter typically runs 1.5–2.5 pages (single-spaced, professional letterhead) and follows this structure:

1. Recommender's credentials: Who they are, institutional affiliation, how long they've been in the field, their own recognition and expertise (publications, grants, positions held). This establishes why their assessment counts.
2. How they know the applicant's work: Specific mechanism (read their papers, cited their methods, attended their talks, reviewed their grants). This establishes independence or explains the relationship context for dependent recommenders.
3. Assessment of specific work: Three to five specific contributions, with enough detail to show the recommender actually read the papers. Vague praise is weak; specific analysis of specific papers is strong.
4. Prong-targeted assessment: Why the work has national importance (prong 1), why the applicant specifically is well-positioned (prong 2), and optionally why the U.S. benefits from this applicant remaining (prong 3).
5. Endorsement: A direct statement that the recommender believes approval is warranted, with the prong-3 argument if applicable.

Common kill-shots

• Template letters: Five letters that are clearly the same template with names swapped. USCIS adjudicators read hundreds of these and recognize templates immediately. Each letter should be unique in structure, voice, and specific examples.
• Vague prong-1 statements: "AI is a critical field for the U.S." is not prong-1 support. It needs specificity: which federal priority document, which specific application, which documented gap.
• No specificity on the applicant's work: A letter from a prominent scientist that doesn't cite a single specific paper by the applicant reads as a favor, not an assessment. USCIS discounts it accordingly.
• Missing the proposed endeavor: Letters that discuss the applicant's past work but don't connect it to the specific proposed endeavor in the petition leave prong 2 incomplete.
• All from the same institution: Five letters from the same university, especially the same department, is a weak independence argument even if all five are technically independent (never co-authored).

If your petition has already received an RFE challenging your letters, see how to respond to an NIW RFE. The EB-2 NIW guide covers the full evidence framework.

Not legal advice. Consult a qualified U.S. immigration attorney before filing.