About Salary Data
Our Editorial Team
Salary Data is produced by the H-1B Salary Data Editorial Team, a small independent data team that builds reference sites from public government datasets. We are not affiliated with the U.S. Department of Labor, with any employer named on this site, or with any immigration law firm or consultancy. Our work on this site consists of downloading the Department's published disclosure files, cleaning them under published rules, computing medians and distributions, and writing the explanations that surround the tables. When we make an editorial judgment — such as the outlier rule or the decision to headline annual-wage medians — we document it on the methodology page rather than burying it in code.
Methodology
Data: U.S. Department of Labor OFLC LCA disclosure files FY2024–FY2025, retrieved October 2026, from the Department's foreign labor performance data page at dol.gov. We used five quarterly LCA disclosure files — FY2024 Q4 and FY2025 Q1, Q2, Q3 and Q4 — which together cover the last quarter of FY2024 and the full FY2025 fiscal year. From 717,449 total rows we kept 662,198 cases whose CASE_STATUS is exactly "Certified"; denied, withdrawn and certified-withdrawn cases are excluded. PERM and H-2A disclosure files are excluded entirely. Headline medians use filings whose WAGE_UNIT_OF_PAY is Year; other units are counted in case totals but not converted into headline medians. Wages below $15,080 or above $1,000,000 a year are excluded from wage statistics as likely placeholders, and SOC 15-1252 is flagged throughout as a reclassification catch-all. The full methodology, with worked explanations, is on the methodology page.
Employers appear under the name recorded in the disclosure files, normalised only for capitalisation and spacing when grouping spelling variants of the same filer. Occupations are the Standard Occupational Classification codes and titles recorded on the filings. Worksites are the city and state of the place of employment as recorded, not employer headquarters. No figure on this site is estimated, modelled or taken from a survey: every number is a count, median, mean or percentile computed by our script from the downloaded rows, and the site manifest records the row counts and cleaning totals behind the build.
What We Publish
This site publishes three kinds of pages. Data pages present computed distributions for a single employer, occupation, state or city: a headline median, a mean, the middle-half range between the 25th and 75th percentiles, filing counts, and the rankings that explain what sits behind the headline. Hub pages collect those data pages into ranked indexes so readers can move from a national picture to a specific segment in one click. Guide pages explain the system that produced the data — what an LCA is, how prevailing wage levels work, why one occupation code acts as a catch-all, and how an offered wage differs from actual pay — because the tables are only as trustworthy as the reader's understanding of the form behind them.
Coverage and Updates
The current edition covers five quarters of filings: FY2024 Q4 and the four quarters of FY2025, retrieved from the Department of Labor in October 2026. We chose a five-quarter window deliberately: it is long enough that a single quarter's hiring spike does not dominate an employer's median, and recent enough that the figures describe the current labour market rather than a historical one. When we refresh the site with newer disclosure files, we will rebuild every page from the new files under the same rules, update the last-updated line, and record the new retrieval date here and in the methodology. Figures from different editions should not be mixed without noting the period each one covers.
Independence
No employer named on this site has paid for placement, ranking or coverage, and no figure here can be changed by request from the organisation it describes. Rankings are mechanical: employers are ranked by certified case count, occupations by certified case count, and places by certified case count, all computed by the same script from the same files. If an organisation believes its filings have been grouped under the wrong name, that is a data question we will review on the evidence of the files themselves, as described on the contact page — it is not a commercial negotiation.
We correct errors when they are pointed out with evidence. If a figure on this site does not match a computation from the same public files under the rules stated here, contact us through the contact page with the page address and your working, and we will either fix the figure or explain the difference. Corrections are recorded by updating the page's data and its last-updated line.