The officer reads the standard first. The criteria are only how you prove it.

The order the reading happens in. Every step is sourced in the list at the foot of this page.
The stage	What the officer is doing there
The statute, read first	INA 101(a)(15)(O)(i): extraordinary ability in the sciences, education, business or athletics, demonstrated by sustained national or international acclaim, with achievements recognised in the field through extensive documentation. One sentence. No number in it.
The height, fixed before any exhibit is opened	8 CFR 214.2(o)(3)(ii): a level of expertise indicating the person is one of the small percentage who have arisen to the very top of the field of endeavor.
The criterion, which is a route and not the test	8 CFR 214.2(o)(3)(iii): a major internationally recognised award such as a Nobel Prize, or at least three of eight listed forms of documentation, or comparable evidence where the criteria do not readily apply.
Fit, checked criterion by criterion	Does the exhibit match the criterion's own words? For criteria carrying qualitative requirements, the quality and calibre of the evidence is assessed here, not later.
The comparison group, fixed at this point	The field of endeavor the person claims. The top of a field is a position relative to other people in it, so the boundary of the field decides whether the standard is reachable at all.
The whole record, read at once	Every exhibit in the file, including material that fits no criterion, weighed together against the standard. Some evidence weighs more alone; other evidence only works in company.
What the officer may not do	Decide the small-percentage question at the criterion stage. Deny because a particular evidence type is absent when other evidence meets the criteria.
What comes out of it	A decision on the type and quality of evidence submitted, rather than on assumptions about evidence that was not submitted.

The short version

The O-1A criteria are not the test. They are the permitted way of proving the test, and the two get read in that order. An officer holds the standard for the whole read, checks the exhibits against the criteria as they arrive, and then decides once, on the whole record, whether the person in front of them is one of the small percentage who have arisen to the very top of the field of endeavor.

The standard is one sentence long. Under section 101(a)(15)(O) of the Act a person qualifies as an O-1A if they have extraordinary ability in the sciences, education, business or athletics, demonstrated by sustained national or international acclaim, with achievements recognised in the field through extensive documentation, and they are coming to continue work in that area. The regulation sets the height: extraordinary ability in these fields means a level of expertise indicating that the person is one of the small percentage who have arisen to the very top of the field of endeavor. There is no number in that sentence. There is no score to reach and no partial credit.

The criteria sit underneath, at 8 CFR 214.2(o)(3)(iii). The petition has to show a major, internationally recognised award such as the level of a Nobel Prize, or at least three of eight listed forms of documentation, or comparable evidence where the criteria do not readily apply to the occupation. Three is a floor, not a score. It is the minimum quantity of evidence that permits the officer to move to the substantive question, and it does not answer it.

That is why a file can clear three criteria and still be denied, and why that outcome is not the officer misreading anything. USCIS states the principle in terms on the immigrant side of the same standard: objectively meeting the regulatory criteria in the first step alone does not establish that the person in fact meets the requirements for classification. Its own illustrations are the giveaway, because it uses the two criteria that are easiest to prove. Participating in the judging of the work of others in the same or an allied field, alone, should satisfy the criterion regardless of the circumstances. Publishing scholarly articles, alone, regardless of calibre, should satisfy the criterion. In both cases the next step requires the officer to evaluate that activity to determine whether it indicates the person is one of the small percentage at the very top of the field, enjoying sustained national or international acclaim.

Read the criteria as the exhibits list, then, and read the standard as the argument. A file that presents three criteria has done the inventory. A file that presents three criteria and then shows, with evidence, where the person sits inside a defined field, has done the case.

The standard is read first, and the regulation orders the inquiry that way

8 CFR 214.2(o)(3)(i) does not begin with the eight criteria. Extraordinary ability in the sciences, education, business or athletics must be established for an individual alien. An O-1 petition must be accompanied by evidence that the work the person is coming to the United States to continue is in the area of extraordinary ability, and that the person meets the criteria in paragraph (o)(3)(iii) or (iv). Three requirements, in that order in the regulation's own sentence: one person, one area of ability, then the criteria.

Two things follow from the wording. The classification is personal, so an exhibit that proves an institution's distinction proves nothing about the person until the file connects the two. And the work to be performed has to sit inside the area of extraordinary ability, which means the officer reads the future employment and the past record against the same field definition. A record in one field and a job in another is a mismatch that no amount of quality in either half repairs.

The order also has a practical effect on how long the officer spends on each question. The criteria are mechanical by design: each criterion is a sentence describing a type of evidence, and checking fit is comparison work. The standard is a judgment, and it is the judgment the decision has to survive review on. Officers therefore read the criteria quickly and read the standard slowly. A file that supplies only the fast half has left the slow half to be written by the person deciding the case.

There is one piece of the record that is read even earlier than the regulation: the petition itself, and who filed it. Under 8 CFR 214.2(o)(1)(i), a qualified alien may be authorised to come to the United States to perform services relating to an event or events if petitioned for by an employer. There is no self-petitioned O-1A. The employer or agent is the petitioner, carries the burden, and is the entity an RFE is addressed to. That is a structural fact about this category, not a formality, and it is the first thing a file reveals about itself.

Meeting a criterion is not satisfying the standard

The regulation's own grammar says the criteria are evidence of the standard

The operative sentence at 8 CFR 214.2(o)(3)(iii) is that an O-1A alien must demonstrate sustained national or international acclaim and recognition for achievements in the field of expertise by providing evidence of the listed items. The acclaim is the thing to be demonstrated. The eight items are the evidence by which it is demonstrated. Swap those two roles and you get a checklist case, which is the most common shape of a denial that surprises the applicant.

The two-step structure separates fit from standing, and step one is deliberately narrow

USCIS runs this standard through a two-step analysis. Step one determines, by a preponderance of the evidence, which evidence objectively meets the parameters of the regulatory description for that type of evidence. Step two is the final merits determination, where all the evidence is evaluated together in the context of the high level of expertise the classification requires. The instruction for step one is explicit that an officer should not yet decide whether the person is one of that small percentage who have risen to the very top of the field. That is not a licence to skip the question. It is a scheduling rule: it tells you where the question gets answered, which is at the end, on everything.

Quality is assessed inside the criterion, not only afterwards

One sentence in the guidance does more work than any other for people preparing a file: for the purposes of step one, officers should consider the quality and calibre of the evidence in determining whether a particular regulatory criterion has been met, to the extent the criterion has qualitative requirements. So the soft half of the assessment is not postponed. A criterion that says major significance, or nationally or internationally recognised, or distinguished reputation carries a quality condition inside it, and the officer applies it at the point the exhibit is read.

Two prohibitions protect a well-built record

The officer may not decide the small-percentage question at the criterion stage, and may not deny a petition merely because a particular type of evidence is absent when other evidence meets the criteria. USCIS gives the worked version of that second rule: an officer may think that if a person is extraordinary there should be published articles about them, but cannot deny the petition because no published articles were submitted, so long as the petitioner has submitted other evidence meeting the qualifying criteria that demonstrates the person is in fact extraordinary. Approval or denial turns on the type and quality of evidence submitted, not on assumptions about what was not submitted.

This is the same logic on the immigrant side, which matters if you file both

The extraordinary ability green card at 8 CFR 204.5(h) uses the identical regulatory definition, and EB-1A carries an express two-step analysis with the same prohibition and the same final merits step. The O-1A-specific material sits in the USCIS Policy Manual, Volume 2 Part M Chapter 4, with an appendix titled Satisfying the O-1A Evidentiary Requirements, which the agency's own O-1 page points readers to for how it evaluates evidence. A person preparing both filings is preparing for the same reading twice, and an O-1A built to the standard rather than to the criteria list is most of the EB-1A argument already assembled.

What the officer weighs once the record is read as a whole

Recognition is expected to travel further than the person's own circle

The guidance states that it is generally expected that one whose accomplishments have garnered sustained national or international acclaim would have received recognition for their accomplishments well beyond the circle of their personal and professional acquaintances. That single expectation disposes of a whole class of files: the record assembled entirely from people the person has worked with, praising work those people were paid to be part of.

Letters carry weight according to their basis of knowledge

Letters of support, in the agency's words, are not without weight but should not form the cornerstone of the claim; the statements in them should be corroborated by documentary evidence in the record. Letters that merely reiterate the definitions of the classification, or make general and expansive statements about the person and their accomplishments, are generally not persuasive. The relationship or affiliation between the person and the witness is a factor. Some letters, read in context, do no more than establish that the person is a competent, respected figure in the field, and the record around them lacks the concrete evidence that would turn that into standing.

Standing data is read as a distribution, not a total

When the record reaches this stage, the considerations USCIS lists are comparative by construction: publication in journals of high rank relative to others in the field, with the journal's own measure of standing supplied by the petitioner; authorship position, where senior or sole authorship carries particular weight; the total rate of citation to the person's body of work as high relative to others in the field, for example a high h-index for that field; and documented employment or research experience at leading institutions in the field. Every one of those asks the same question in a different currency: where does this person sit relative to the field.

Evidence that fits no criterion still counts, and evidence that fits one perfectly can still be thin

At the final merits step the officer considers any potentially relevant evidence in the record, even evidence that does not fit one of the regulatory criteria or was not presented as comparable evidence. The instruction cuts in both directions. Some evidence weighs more favourably on its own, while other evidence is more persuasive when viewed with other evidence. A single strong exhibit sitting in a record that gives it no company gets read exactly as it is: one strong exhibit.

Where the reading turns against a strong file

The field is drawn so that the top of it cannot be located

The standard is relative, so the field is the comparison group. Draw it as an employer's department and there is no national or international field to have risen to the top of. Draw it as an industry and nobody is at the top of it. The work of fixing the field happens before filing, and it has to survive the record the officer can see, because the citation data, the venue standing and the letters all get evaluated against the field the petition claims.

The acclaim is real but historical

The statutory word is sustained, applied to acclaim, and it is doing work. A record whose strongest evidence sits years behind the claimed area of ability raises a question the file has to answer rather than avoid: whether the person still holds the level of acclaim they were once recognised for, in the field they are now coming to work in.

The exhibits establish the organisation rather than the person

The critical-or-essential-capacity criterion at 8 CFR 214.2(o)(3)(iii)(B)(7) is about the alien having been employed in a critical or essential capacity for organizations and establishments that have a distinguished reputation. The organisation's reputation is a condition of the criterion, so evidence of it is necessary. It is not sufficient, because the criterion is about the person's capacity within it. A file that proves the lab is famous and leaves the person's function implicit has proved the condition and not the criterion.

A previous approval is offered as the argument

A prior O-1 approval is a relevant consideration and can be an indicator of eligibility, and it is not determinative. Some courts have asked why a person previously classified in a roughly analogous nonimmigrant category was later found ineligible for the immigrant classification, which is why officers issuing denials in those cases should, where possible, briefly explain why the earlier approval does not carry the later petition. For the petitioner the lesson is narrow and useful: the earlier approval is a fact in the record, never the conclusion, and a file that leans on it has stopped arguing.

Three criteria, one number

This is an illustration of the reading, not a report of a case. A machine-learning engineer, eight years in, employed by a company that is filing the O-1A petition. The file claims three criteria and can prove each of them.

Judging: two programme committees for named conferences in named years, with the invitation letters and the published committee listings. Scholarly authorship: eleven papers in venues whose standing is established from the venues' own circulation and acceptance information, plus a citation record given as a distribution across the person's work rather than a single total. High salary: the employment contract, an offer letter, and a compensation benchmark covering the same occupation, the same metropolitan area and the same currency, with the survey's method attached.

Step one is not close. The officer checks three exhibits against three criterion sentences and the fit is exact. The file then has to survive the second read, and this is where the outcome is decided. The claimed field is narrow and defensible: ranking and recommendation systems for subscription marketplaces, a body of knowledge other people work in. The officer asks where this person sits in it. The record answers with the two programme committees, which are evidence of participation rather than prominence, and a citation distribution that is respectable for eight years of work without being unusual in that field.

What the denial is not about is the criteria. The engineer met three of them. What the record never establishes is the sentence the regulation put first: that this is one of the small percentage who have arisen to the very top of that field. Two additions would have changed the read. Independent corroboration of adoption — named companies that built on the ranking method, evidenced by those companies' own engineering publications or product documentation rather than by a letter asserting it. And three letters from researchers in the same field who have never worked with the person, each explaining what they know the field to contain and where this work sits in it relative to the rest.

The arithmetic in this illustration is worth stating plainly, because it is where the high-salary criterion is most often misread. The O-1A text asks for evidence that the person has commanded a high salary or other remuneration for services, evidenced by contracts or other reliable evidence. A contract proves the salary. It does not prove that the salary is high, and it does not establish the comparison. The number is only evidence of standing once the file shows what the number is being compared against.

Questions this raises

How does USCIS actually read the O-1A criteria?

The officer holds the standard first — a level of expertise indicating the person is one of the small percentage who have arisen to the very top of the field of endeavor — then checks each exhibit against the criterion it was submitted under, and finally evaluates the whole record together. Meeting three criteria is the threshold, not eligibility.

Can I get an O-1A with exactly three criteria?

Three criteria is the minimum the regulation allows where there is no qualifying major award, so three is legally sufficient at the criterion stage. It is not sufficient on its own for approval, because the record still has to establish sustained national or international acclaim at the top of the claimed field.

What does 'one of the small percentage at the very top of the field' mean in practice?

It means a relative position, judged against evidence of where other people in the same field sit: citation rates for that field, venue standing, authorship position, institutional calibre, and recognition from people outside the applicant's own professional circle. There is no fixed benchmark stated in the regulation.

Who decides what my field of endeavor is?

The petition states the field and the officer assesses the record against it. The definition therefore has to be defensible in both directions: narrow enough to be a field with a top, and broad enough to contain the community whose recognition the evidence comes from. That choice is made before filing.

Does a previous O-1A approval help an EB-1A petition?

It is a relevant consideration and can indicate eligibility, but it is not determinative. Where a person previously classified as an O-1 is later found ineligible for the immigrant classification, officers issuing the denial should explain why the earlier approval does not carry the later petition.

Why was my O-1A denied when I met the criteria?

Usually because the record proved three criteria and never argued the standard. The criteria are evidence of sustained national or international acclaim, not the acclaim itself. A file built as a checklist leaves the officer to assemble the whole-record case unassisted, and that is the case the decision turns on.