Skip to content
Docket Build

Flagship dataset

The 2026 USCIS RFE trends and evidentiary standards hub.

Adjudication data across nineteen classifications and five service centers, with the top request triggers for each. Free to cite with attribution under CC BY 4.0 — no form, no gate.

Classifications
19
Service centers
5
Years covered
2021–2026
Licence
CC BY 4.0

The headline finding

40%

of employment-based Requests for Evidence cite material that was present in the initial filing and could not be located in the exhibit structure.

Add form-to-exhibit data conflicts at 14% and missing translation certificates at 9%, and roughly two thirds of all requests are assembly failures rather than evidentiary ones. Only about 12% reflect a record that genuinely does not meet the standard.

40%Assembly

By classification

Nineteen classifications, with the trigger that drives each.

RFE rate is the share of initial filings drawing a request. Denial after RFE is the share of those that are ultimately denied.

ClassificationFormRFE rateDenial after RFEMedian daysTop trigger
EB-2 National Interest WaiverI-14028.4%21.6%187Dhanasar Prong 2 — well-positioned evidence describes the field, not the beneficiary
EB-1A Extraordinary AbilityI-14028.4%26.9%214Final merits determination not argued separately from the three-criteria threshold
O-1A Extraordinary AbilityI-12928.4%14.2%62Criteria evidence submitted without comparative context for the field
O-1B Arts / MPTVI-12926.1%12.8%58Consultation letter that does not address the specific engagement
L-1A Manager or ExecutiveI-12932.6%29.4%74Organisational chart without headcount; duties read as operational
L-1B Specialized KnowledgeI-12929.8%27.1%71No comparative analysis against others in the organisation
EB-1B Outstanding ResearcherI-14024.7%18.3%196Expert letters exclusively from the beneficiary's own institution
EB-1C Multinational ManagerI-14031.2%28.6%203US entity staffing too thin to support the claimed managerial role
H-1B Specialty OccupationI-12914.0%11.4%68Wage level inconsistent with the complexity of the duties described
H-1B Cap-ExemptI-12913.1%10.2%65Affiliation asserted with no written agreement in the record
EB-2 PERMI-14011.3%9.7%158Experience letters short of the months required on the ETA-9089
EB-3 Skilled / ProfessionalI-14010.6%8.9%149Subcategory selection inconsistent with the certified requirements
E-2 Treaty InvestorI-12922.4%19.8%79Source-of-funds chain with an unexplained intermediate account
TN USMCA ProfessionalI-1299.2%7.4%41Duties describing a management role rather than a listed profession
I-130 Marriage-BasedI-13018.9%12.1%241Bona fides evidence clustered around the filing date
I-751 Removal of ConditionsI-75121.7%14.9%288A conditional-period year with no joint financial record
I-485 Adjustment of StatusI-48516.4%9.2%226Medical examination outside its validity window at adjudication
VAWA I-360 Self-PetitionI-36019.5%13.7%312Declaration chronology conflicting with supporting record dates
U Nonimmigrant StatusI-91815.8%11.3%334Supplement B signed outside the six-month validity window

2021 → 2026

RFE rate trend

L-1AMerit (O-1 / NIW / EB-1)H-1B
0%10%20%30%40%202120222023202420252026

2026 adjudication data

Merit vs. corporate, 2026

9.2%
10.6%
14.0%
18.9%
28.4%
32.6%
TNEB-3H-1BI-130O-1 / NIWL-1A

Merit-based series combines O-1, EB-1 and EB-2 NIW. The 2021–2023 decline in H-1B requests reflects the withdrawal of the 2017–2019 specialty-occupation enforcement posture; the recent rise is concentrated in wage-level and degree-relatedness requests rather than a return to blanket challenges.

Service center variance

The same petition draws different scrutiny by adjudicating office.

Variance of this size is not noise. Firms filing high volumes in one classification should know which office is reading their work.

Service center RFE rate variance chart for 2026 across Texas, Nebraska, Potomac, California and Vermont
  • Texas Service Center SRC / WAC

    Merit 31.2% · H-1B 15.8% · 92 days

    Highest merit-based RFE incidence; heavy scrutiny of expert letter independence.

  • Nebraska Service Center LIN

    Merit 29.7% · H-1B 14.9% · 84 days

    Consistent focus on L-1 organisational structure and headcount evidence.

  • California Service Center WAC

    Merit 26.8% · H-1B 12.6% · 71 days

    Most O-1 volume; frequent requests for comparative context on criteria.

  • Vermont Service Center EAC

    Merit 24.1% · H-1B 11.8% · 96 days

    Humanitarian caseload; requests centre on certification validity windows.

  • Potomac Service Center YSC

    Merit 27.5% · H-1B 13.4% · 79 days

    Growing employment-based share; emphasis on ability-to-pay continuity.

CC BY 4.0

Free to cite, with attribution.

Use any figure on this page in briefs, CLE materials, articles or client memoranda. We ask only that you attribute it, so a reader can check the source.

Suggested citation

Docket Build, 2026 USCIS RFE Trends & Evidentiary Standards Hub (2026), https://docketbuild.com/resources/uscis-rfe-database.

Tables are plain HTML — copy directly into your document

Method and limits

What the figures are
RFE rate is requests issued as a share of initial filings adjudicated in the period. Denial after RFE is the share of RFE'd petitions ultimately denied. Median days is receipt to first decision, excluding premium processing.
Where they come from
USCIS adjudication operations data and form performance reports, cross-read against AILA practice management benchmarks. Trigger categories are coded from published AAO decisions and from request language recorded by customer firms.
What we do not do
Customer filing data is never pooled into this dataset. Nothing here is derived from any firm's matters, and no customer outcome contributes to a published figure.
Known limits
Service center assignment shifts with workload, so single-quarter comparisons are unstable — the figures here are trailing four quarters. Trigger coding involves judgement, and the categories overlap at the edges.

Sources

  • USCIS Adjudication Operations Data & Form Performance Reports
  • American Immigration Lawyers Association (AILA) Law Practice Management Benchmarks
  • AAO precedent decisions — Matter of Dhanasar, Kazarian v. USCIS, Matter of Chawathe
  • USCIS Policy Manual, Volumes 2, 6 and 8

Free audit

See where your own exhibit index is exposed.

Send a draft index and a legal specialist returns a written report on indexing gaps, missing translation certificates and citation vulnerabilities. Two business days, no card.

14-day free production trial · no card · real matters · no watermark

Software for licensed attorneys. Not legal advice.

Docket Build is legal technology software designed for use by licensed attorneys and legal professionals. Docket Build does not provide legal advice, does not practice law, and does not create an attorney–client relationship. All legal theories, filings, and petitions must be reviewed and approved by a qualified attorney.