How much must you earn for the high-salary criterion?

What a correct comparison is made of

The notices list what the officer weighs when comparing your income to survey data: the description of the occupation, the validity of the survey itself, location and currency, and the salary rate being measured. Four independent axes, and missing on any one of them voids the comparison entirely.

Occupation. The data has to match your role, not the industry at large. In one of the cases we parsed the officer noted outright that the salary data submitted was not comparable to the petitioner's position. A generic 'IT salaries' line against a specific role is the standard miss.

Location. The requirement is to show you earn significantly more than people in similar positions in the same geographical area and industry. A national average against a capital-region income does not survive the comparison.

Currency and period. The data is taken in the currency of payment and for the same years. The agency does not accept conversion into dollars - and this is not a formality: purchasing power and pay levels differ in ways an exchange rate does not capture.

The rate being measured. An annual figure is compared with an annual one, an hourly rate with an hourly rate. Salary, bonuses, dividends and money arriving in an account cannot be mixed: in one notice the officer analyses separately the difference between a bank balance and actual remuneration.

Which data sources work

For US employment the base set is clear: official wage statistics by occupation and area, industry compensation surveys, W-2 or 1099 forms for the high-income years. Department of Labor data is useful as a reference point, but on its own it is not enough - the notices carry a specific caveat about this.

For employment outside the US: official national occupational statistics, industry salary surveys in local currency, and your country's tax filings establishing the annual income. It matters to explain what the source is: the officer is not obliged to know what your national statistics office does.

An employer's justification also works - an explanation of why the company pays above market. It is named explicitly among the possible forms of evidence, and for non-standard roles it is often the only way to show comparability.

What does not work: screenshots of job ads, salary aggregators with no methodology, your own assessment of the market, conversion at an exchange rate. The officer reads all of it as assertion without support - and assertion without support does not meet the burden of proof.

Why the acceptance rate on this criterion is so low

14% is among the lowest rates in our set - only original contribution (4%), membership (12%) and commercial success sit below it. The reason is not that petitioners earn little; it is that they prove the wrong thing.

The typical structure of a failing section: contracts, payslips, statements - and not one comparison table. The petitioner shows the income is high by their own standards, while the criterion asks for a position relative to others.

The second most common failure is a non-comparable basis. The data is there but for the wrong occupation, the wrong country, the wrong years or in a different currency. Formally documents were attached; substantively no comparison happened.

One more detail visible in the notices: this is the only criterion where, in our set, Matter of E-M- on the standard of proof is cited six times - the decision saying that truth is determined by the quality of the evidence, not its quantity. It is precisely the criterion where volume most often arrives in place of a comparable basis.

The practical conclusion: if you have no comparable data, it is more honest not to claim this criterion at all and to move the effort to one where the evidence genuinely exists. A weak point adds no weight but gives the officer grounds for an extended objection.