| Criterion | What an academic already has | What still needs building |
|---|---|---|
| Authorship of scholarly articles | Journal papers and conference proceedings | Evidence of venue standing, and a byline note for large collaborations |
| Judging the work of others | Reviewer reports, grant panels, programme committees | Letters from editors confirming role, period and volume |
| Original contributions of major significance | A body of work other groups build on | A citation map showing who used it and what changed |
| Published material about you | Interviews and news coverage of results | Coverage where you are the subject, not an institutional release |
| Membership in selective associations | Society membership, some fellowships | Election judged by recognised experts on achievement |
| High remuneration | Salary set by a university or institute scale | Benchmark data for the field and the country, not one figure |
The short version: your publication record is already most of the file
You have a publication list, a review history and results other groups build on. The O-1A is the extraordinary ability subcategory for the sciences, education, business and athletics. It is closer to your record than most researchers expect.
The standard is narrow. It asks whether you are one of the small percentage who have arisen to the very top of the field of endeavor, which is a question about standing rather than volume.
You need a major internationally recognised award, or at least 3 of the 8 listed criteria. Three is the legal floor, and a file built to exactly 3 reads as a file that stopped at the minimum.
Three of those criteria are native to research: authorship, judging and original contributions. That is why a working scientist often has the substance before anyone explains the form.
Most academic files are strong on authorship and weak on adoption, because a publication list shows output while the criteria want consequence. That gap is fixable in a month.
The problem is that a CV (curriculum vitae) lists while a petition argues, so the same career arrives as a list of titles rather than as evidence that the field moved. Format loses cases substance should win.
What follows is how citations, peer review and authorship are actually read, and where your file usually goes wrong. It is written for the researcher who has the record and not the argument.
How citations are read, and what a count misses
There is no citation number in the regulation, because the criterion is original contributions of major significance. A count is not the test.
What officers look for is who cited your work, what they did with it and whether anything changed outside the group that produced it, because adoption is the word that matters.
A method another lab runs, a dataset a product depends on, a result written into a standard, a licence a company took. Those are contributions with a trace.
Your citation volume without that trace is weaker than it looks. Self-citation, citation rings and citations from the same 3 labs get discounted. Reviewers notice the pattern quickly.
One citation of your paper in a technical standard or a textbook chapter often carries more weight than 50 from collaborators. Rarity is not the point; consequence is.
Field differences matter too, since your field's citation rate is not another field's. A raw number compared across disciplines tells an officer nothing useful.
Letters turn citations into an argument, and they only carry weight when the writer has no stake in your career, which usually means someone at another institution who used your method. A famous supervisor is not that person.
Choose 3 or 5 of your papers rather than 40, and build a short citation map for each. Who used the result, what they built, and what an independent expert says it changed.
That is a document, not a list, and it answers the question the criterion asks: whether your work mattered to people who owed you nothing.
How peer review is read as judging
Judging is the criterion most researchers already meet without noticing, and it covers the reviewing you already do: manuscripts, grant panels, programme committees and doctoral work.
Payment is irrelevant. What has to exist is documentation: editor letters, panel invitations, reviewer records and the years in which the work happened.
A claim that you reviewed 60 papers for 6 journals is worth little on its own, since an officer cannot check it and cannot see whether the journal selects its reviewers. Letters fix that.
Ask editors now, while they remember you. One paragraph confirming that you reviewed for the journal, how often and over which period carries the criterion.
A useful letter says what you reviewed, for which journal or panel, over what period and how often. It does not need to argue about your standing, and asking an editor to do that is how requests get declined.
Being reviewed is not judging, and publishing in a peer-reviewed journal is not judging either. Files that conflate the two waste a page of your evidence and weaken the rest.
An editorial board seat is usually argued as judging. Editors decide what enters the literature. It is stronger than a single review, because the role is public and ongoing.
Depth beats volume here. Two years of steady reviewing reads better than 40 reports in one month, which looks like a favour rather than a role.
Authorship, press, and the difference between 2 criteria
Authorship of scholarly articles and published material about you are separate criteria, and researchers merge them by accident. One is work you wrote; the other is work someone wrote about you.
For your authorship evidence, position and venue carry most of the reading, since a first author on a paper in a recognised journal is doing different work from the ninth name on a 30-author paper. Officers notice.
A preprint is a manuscript rather than a vetted publication, so while it can support a contribution argument the authorship criterion reads more clearly once the paper is accepted.
Venue quality gets checked, since one paper in a title reviewers recognise is worth more than 5 in venues nobody has heard of, and a journal that accepts anything damages your file.
Published material about you works differently, since a profile, an interview or a news story about the result that names you is coverage of a person, which is what the criterion wants.
A university press release is written by the institution and reads as institutional promotion, so although it quotes you it is not independent coverage of your work. Officers treat it that way.
Contribution statements help here, since many journals now publish who did what. Attach the statement rather than describing the division of labour yourself.
Both criteria reward one habit: keep the link between your work and your name visible. Co-authored work needs a letter or a byline note explaining what you did.
What does not count, and why officers discount it
Your h-index is a summary rather than evidence, computed from the same citations an officer can already see, and it hides what other people did with your results.
Conference abstracts, posters and extended abstracts are not articles. They announce work in progress. A file padded with them reads as a file short of publications.
Invited talks at your own institution do not count as recognition. A departmental seminar is a courtesy. A seminar at another university is a small one unless the invitation came with standing.
Internal prizes fail for the same reason as internal awards anywhere. A teaching award, a travel grant or a poster prize is recognition from your employer, and the criteria ask about the field.
Society membership is worth nothing when anyone at your career stage can join by paying dues, and honorary memberships handed to a whole cohort have the same problem. Selection has to mean something.
Coverage of the institution is not coverage of you, since a story about your university's research that quotes your supervisor says nothing about your standing, however large the outlet.
The pattern is consistent across all of it. Anything that arrives automatically, or that you or your employer produced, is not evidence that the field chose you.
O-1A or EB-1A: one record, 2 different questions
The O-1A is temporary work authorization tied to a petitioner, while an EB-1A, the extraordinary ability green card petition, is a separate petition under 8 CFR 204.5(h) with its own criteria. They draw on the same record and ask different questions.
O-1 requires a petitioner, an employer or an agent, there is no self-petition route, and a written advisory consultation from a peer group or person of distinction must accompany the petition.
What O-1 gives you is time: an approved petition is valid for up to 3 years, and extensions come in 1-year increments. That is a runway, not a queue.
The classification also sits outside the numerical caps that apply to H-1B and H-2B, the 2 main temporary work categories. Caps that bind others do not bind this route.
Time on O-1 does not create green card queue position, so years spent on it accumulate toward nothing. The green card route is a separate filing built on its own evidence.
That difference changes the timing. An O-1 can be extended while the record grows, and a green card petition is decided on the record as it stands. One buys time; the other ends the question.
If you need to start a job this year, the O-1A is the instrument that exists. If your record is complete and you want to stop renewing, the immigrant petition settles status.
Two steps apply to both. First whether the criteria are met, then a final merits determination on the whole record, which is where a thin but technically compliant file fails.
| Question | O-1A | EB-1A |
|---|---|---|
| What it gives you | Temporary permission to work for a petitioner | Lawful permanent residence, if approved |
| Who files | An employer or an agent; no self-petition | A separate petition under 8 CFR 204.5(h) |
| Where the criteria sit | The 8 criteria at 8 CFR 214.2(o)(3)(iii) | The extraordinary ability criteria at 8 CFR 204.5(h) |
| How long it lasts | Up to 3 years, then 1-year extensions | Permanent status, once granted |
| What O-1 time does for it | Nothing; no queue position is created | It is the petition that produces the green card |
Questions this raises
How many citations do I need for an O-1A?
There is no number. The criterion is original contributions of major significance, and citations are one way to show adoption. Who cited the work and what changed as a result matters more than a total, and a total without field context proves little.
Does unpaid peer review count?
Yes. The criterion covers judging the work of others, and payment for the judging is irrelevant. What matters is documentation: an editor's letter confirming the journal, the role and the period, or panel invitations that name you. Unpaid review is the normal arrangement, not a weakness.
Can a postdoc get an O-1A?
Job title is not the test. The standard is standing in the field, so a postdoc with independent citations, review work and 3 documented criteria can meet it. The petition still needs a petitioner, which can be an employer or an agent.
Is a preprint enough for the authorship criterion?
Usually not on its own. A preprint is a manuscript that has not been vetted, and the criterion asks about articles in professional journals or other major media. An accepted paper at a recognised venue reads far more clearly. A preprint can still support a contribution argument.
Do I need a consultation letter?
Yes. A written advisory consultation from a peer group or a person of distinction is required with the petition. It is separate from the expert letters you collect, and it should come from people who know the field rather than from close colleagues.
Can O-1A time count toward a green card?
No. Time on O-1 does not create queue position. The EB-1A is a separate petition under 8 CFR 204.5(h) with its own criteria, and it has to be filed and argued on its own record. Plan it as its own project with its own evidence.