Evidence Building
How to Document Salary Above the 90th Percentile When BLS Data Understates Your Market
BLS wage data often underestimates compensation in competitive technology and finance markets, leaving petitioners without an accurate benchmark for the O-1A high salary criterion. Here is how to build a multi-source comparison that accounts for total compensation and reflects the petitioner's actual market.
The high salary criterion and its practical weight
The O-1A high salary criterion — one of eight enumerated criteria under 8 C.F.R. § 214.2(o)(3)(ii) — is often the most straightforward to satisfy on paper and the most frequently contested in practice. A petitioner who earns significantly more than peers in their field has clear documentary evidence of recognition, since employers set compensation based on market value and perceived contribution. In practice, many O-1A petitions include salary as a supporting criterion rather than the primary showing, but when documented rigorously, it can carry substantial weight, particularly for petitioners in technology, finance, and other sectors where compensation levels are high and comparable data is publicly available.
The challenge is that USCIS has never specified a precise percentile benchmark for what qualifies as high salary. The regulatory standard requires that the beneficiary command a high salary or other remuneration for services as compared to others in the field. That comparative framing is both the criterion's strength and its complication: the comparison class must be defined, and the definition has a significant effect on whether any given figure appears high. A salary above the 90th percentile nationally may be at the median in San Francisco's technology sector. Defining the right comparison class and supplying documentation that reflects it accurately is where most high-salary submissions succeed or fail.
BLS Occupational Employment and Wage Statistics data is the most commonly cited comparison source for O-1A high salary evidence. It is freely accessible, covers hundreds of occupations, and reports wage percentiles by metropolitan area and nationally — attributes that make it an appealing evidentiary source. But BLS data has meaningful limitations for petitioners in high-compensation sectors. Its survey methodology lags market conditions by 12 to 18 months, it does not capture equity compensation or signing bonuses, and its occupational classifications do not map cleanly onto the job categories used by technology employers. The result is that BLS data often understates the actual market for senior roles in competitive sectors.
What the regulation actually requires
The regulatory text at 8 C.F.R. § 214.2(o)(3)(ii)(A)(8) describes the high salary criterion as covering a high salary or other remuneration for services in relation to others in the field. The phrase other remuneration for services is important and often underutilized — it explicitly encompasses total compensation beyond base salary, including equity grants, performance bonuses, and other forms of pay that reflect the employer's assessment of the beneficiary's value. A petitioner whose base salary falls short of a threshold but whose total compensation package including equity and bonus places them well above peers may satisfy the criterion on total compensation grounds, provided the non-salary components are adequately documented.
USCIS requires a genuine comparison to others in the field — not a comparison to national averages across all occupations, and not a comparison to a strategically narrow subset chosen to make the petitioner appear highly paid. The comparison class should reflect the actual occupational category and geographic market in which the petitioner works. A petitioner employed as a senior machine learning research scientist at a major technology firm should be compared to others in that specific role, not to all software engineers nationally. Building the right comparison class requires knowing which data sources cover that occupational niche and at what geographic level, since national and metropolitan-level wage statistics diverge substantially in high-cost markets.
The I-129 petition must present the comparison affirmatively — adjudicators are not expected to independently research wage percentiles for the petitioner's occupation. The evidence package should document actual compensation, explain the geographic and occupational scope of the comparison group, present the relevant wage statistics for that group, and show clearly where the petitioner's compensation sits relative to those statistics. A petition that submits a pay stub and a BLS table without explicitly calculating and stating the resulting percentile leaves the adjudicator to do interpretive work that may not favor the petitioner.
Evidence that routinely satisfies the criterion
The most persuasive high salary evidence packages combine multiple data sources to triangulate the petitioner's position in the compensation distribution. BLS OEWS data for the relevant occupational classification and metropolitan area establishes a regulatory-recognized baseline. Employer-published compensation surveys from reputable sources — Levels.fyi for technology roles, or third-party benchmarking reports from firms such as Radford, Mercer, or Willis Towers Watson — add specificity and often better reflect actual technology-sector compensation than BLS classifications allow. When two or three independent sources all place the petitioner above the 90th percentile, the cumulative case is substantially stronger than reliance on any single source.
Documentation of total compensation components is essential when base salary alone does not cross the high-salary threshold. An offer letter or compensation summary from the employer that itemizes base salary, annual bonus target, equity grant details including vesting schedule and grant-date value, and any signing or relocation bonus gives the adjudicator a complete picture of total remuneration. Equity compensation documentation should be accompanied by a current valuation source — the employer's most recent 409A valuation for private companies or closing share price at grant date for publicly traded employers — to convert the equity grant into a comparable dollar value for the comparison.
Letters from independent compensation experts or senior HR professionals at companies in the same sector can add qualitative context that raw statistics cannot provide. A declaration from an experienced compensation consultant explaining why the petitioner's total compensation package reflects extraordinary recognition in the field — why it places the petitioner in a category that only a small fraction of similarly titled professionals commands — translates wage data into the narrative the adjudicator needs. These expert letters are most effective when they are specific about the occupational scope, the geographic market, and the comparison methodology, rather than offering general conclusions unsupported by the referenced data.
Evidence USCIS regularly discounts
BLS national wage statistics for broad occupational categories are the most commonly discounted high salary evidence in technology sector O-1A petitions. The problem is not that BLS data is inaccurate — it is that the categories are often too broad to reflect the petitioner's actual market. A petitioner working as a senior machine learning engineer at a major platform company earns in a compensation distribution materially different from the national BLS category for Software Developers, Quality Assurance Analysts, and Testers, which encompasses a wide range of roles at companies of all sizes and in all markets. Using that broad category as the baseline may show the petitioner is above average, but it will not show they are in the top tier of the relevant market segment.
General offer letter language without supporting data is similarly discounted. A sentence stating that the petitioner's compensation is at the 95th percentile for their level is an assertion, not evidence. USCIS adjudicators expect the petitioner to document the basis for any percentile claim, not simply to assert it. An offer letter that cites a specific compensation survey and attaches the relevant percentile data as an exhibit is substantially more persuasive than one that makes an unsupported percentile assertion. The employer's internal attestation is not an independent third-party verification of market standing, even if the underlying number is accurate.
Comparison to company-internal salary bands without external benchmarking data is another common submission weakness. Some petitioners submit evidence that they sit at the top of their employer's internal band for their level and location. Internal salary bands are proprietary and reflect the employer's compensation philosophy, and USCIS cannot verify that those bands are themselves high relative to the broader field. The comparison the criterion requires is to others in the field as a whole, not to the petitioner's position within a single employer's internal structure. Internal band placement can be submitted as supporting context but should not serve as the primary comparator.
How to present borderline compensation evidence
Petitioners whose total compensation places them in the 80th-to-90th percentile range have a viable strategy for the high salary criterion if they approach the evidence with attention to comparison class selection and supplemental framing. The first step is to identify whether a narrower, more accurate comparison class places the petitioner more clearly in the high-salary tier. A compensation figure at the 83rd percentile nationally for a broad software engineering category may be at the 92nd percentile within the specific metropolitan area at the same level of seniority. Narrowing the comparison class to reflect the petitioner's actual market — with supporting data to justify that narrowing — shifts the percentile picture without misrepresenting the compensation situation.
When total compensation including equity is materially higher than base salary, the framing strategy should lead with total compensation rather than base. This requires calculating the annualized value of equity grants at a defensible valuation, adding that to base salary and bonus, and presenting the aggregate against a total compensation dataset. Platforms that report total compensation rather than base salary alone for technology roles are appropriate references when the petitioner's compensation story is primarily about equity and bonus rather than salary. The evidence package submitted to USCIS should explain the calculation methodology clearly so the adjudicator can follow it without making independent inferential steps.
Expert declarations can frame borderline compensation as reflecting a recognized tier of professional standing even when the percentile number is ambiguous. A declaration from a compensation consultant who can explain that a specific total compensation level in a particular market places the petitioner in a cohort of no more than a small fraction of field practitioners — and that such compensation reflects the employer's judgment that the petitioner brings expertise the market prices at a premium — translates a number into a qualitative statement about standing. The combination of a well-constructed percentile calculation and a thoughtful expert declaration providing that calculation with context is the highest-value submission approach for borderline high salary cases.
Building and auditing the high salary file
A well-organized high salary evidence file has five components: a complete compensation statement from the employer documenting all compensation elements for the relevant year; wage comparison data from at least two independent sources covering the relevant occupational classification and geographic market; a calculation exhibit that explicitly places the petitioner's total compensation on the comparison distribution; any supporting declarations explaining the comparison methodology and its significance; and a cover section in the petition brief that narrates the evidence and states the percentile conclusion. The brief narrative is essential — it should not leave the adjudicator to assemble the argument from the underlying exhibits.
Auditing the file before submission requires checking each component against the common weaknesses adjudicators flag. Is the comparison class narrow enough to reflect the petitioner's actual market, and is that choice of comparison class justified by the data sources selected? Is total compensation fully documented with a defensible valuation methodology for any equity component? Does the calculation exhibit explicitly state the resulting percentile rather than leaving it implicit? Are the data sources current enough to reflect market conditions at the time of filing? Is the expert declaration, if included, specific about the comparison methodology rather than offering only a general conclusion?
If the file passes this audit, it presents the high salary evidence in the form most likely to produce a straightforward approval without an RFE requesting additional documentation. If it does not pass — if any element is missing or weak — the petition is an RFE candidate on this criterion. Given that RFEs add months to adjudication timelines and require submitting additional documentation under pressure, front-loading the evidence development to preempt foreseeable RFE concerns is almost always the more cost-effective strategy. High salary is, in principle, one of the more documentable O-1A criteria — the challenge is not the underlying evidence but the discipline to document it completely and accurately from the start.
What we typically gather for this kind of case
| Document | Where to source | Why it matters |
|---|---|---|
| Expert letters | 5–8 independent recognized experts | Quality and independence beat volume |
| Certified translations | ATA-certified translator | Required for any non-English source document |
| Exhibit cover sheets | Drafted by counsel, one per exhibit | Tells the adjudicator what each piece shows |
| Bibliometric reports | Web of Science / Scopus | Quantifies impact for original-contributions criterion |
What we see go wrong, again and again
- 01Sending exhibits without a one-paragraph framing memo explaining what each shows and why it matters.
- 02Relying on volume over specificity — five well-targeted expert letters beat fifteen generic recommendations.
- 03Skipping certified translations or using AI translation for foreign-language source documents.
See if you qualify
Lando reviews your background against the O-1 visa criteria and tells you honestly where you stand. Free, no commitment.