Insights · Data guide
Groundwork research · Published October 19, 2026 · Data as of September 29, 2026 · 14 min read
A Labor Condition Application, or LCA, is the form an employer files with the Department of Labor (DOL) before it sponsors a worker on an H-1B, H-1B1 or E-3 visa. DOL publishes the decided applications every quarter as Excel files. Our copy holds 661,885 certified applications from 71,614 employers. They cover 15 months of DOL decisions (Jul–Sep 2024, Jul–Sep 2025, Oct 2025–Jun 2026). This guide explains what a row is, what the columns hold, which months a file covers and what the files leave out.
Before an employer files an H-1B petition with U.S. Citizenship and Immigration Services (USCIS) or applies for a visa, it submits an LCA, Form ETA-9035, to DOL's Office of Foreign Labor Certification (OFLC). The form is an attestation: the employer states the job, the occupation, the worksite, the wage it will pay and the prevailing wage for the job, and carries the burden of proving that the information is true. DOL reviews each LCA within seven working days for completeness and obvious errors, and an employer files up to six months before the job starts (DOL's H-1B page).
Three programs share the form: H-1B for specialty occupations, H-1B1 for workers from Chile and Singapore and E-3 for workers from Australia. The files mix all three and name the program in VISA_CLASS. It splits the 661,885 certified applications in our copy into H-1B 644,954, E-3 Australian 12,375, H-1B1 Chile 2,647 and H-1B1 Singapore 1,909.
A row is one application, identified by its case number in CASE_NUMBER. A case number reads I-20P-YYDDD-NNNNNN: I-20 and a program digit, the last two digits of a year, the day of that year and a serial number. The program digit names the visa class: I-200 is H-1B, I-201 is H-1B1 Chile, I-202 is H-1B1 Singapore and I-203 is E-3 Australian. A number such as I-200-25273-123456 carries day 273 of 2025, which is September 30. All 676,973 case numbers in our workbooks follow this pattern. The received date, RECEIVED_DATE, falls on the day a number carries for 93% of the rows that have both, and on the day before for 6.7%.
An application can ask for more than one worker. TOTAL_WORKER_POSITIONS holds the number requested: the 661,885 certified applications in our copy request 1,178,277 positions, and 27,627 of them request more than one. A count of rows counts applications and a count of positions counts requested workers. Neither counts the people who took a job.
An application can also list several worksites. 41% of the certified applications in our copy list two or more in TOTAL_WORKSITE_LOCATIONS. The place and the wage columns in the main workbook describe the first worksite. The others sit in a separate Worksites workbook that DOL posts beside the main one, and an Appendix A workbook covers the applications that attach one.
Six more columns split the requested positions by kind of petition: new employment, continued employment, change in previously approved employment, new concurrent employment, change of employer and amended petition. They hold counts of workers, and they add up to TOTAL_WORKER_POSITIONS on 99.8% of the certified applications.
Requested positions by kind of petition, share of 1,178,277
New employment is 30% of the requested positions across the workbooks: 19% in FY2024 Q4, 19% in FY2025 Q4 and 34% in FY2026 Q3. The rest continue, change or amend an earlier approval, so a row in these files describes a continuing worker more often than a new one. The share moves with the months a workbook covers.
The workbooks have 97, 98 and 98 columns (FY2024 Q4, FY2025 Q4 and FY2026 Q3). These are the ones most readers use, in DOL's wording from the record layout. The rest hold the contact details of the employer, its point of contact, its agent or attorney and the preparer, and the form's other attestations.
Every column listed here is filled on every certified row except PW_WAGE_LEVEL (91%), WORKSITE_COUNTY (86%), WAGE_RATE_OF_PAY_TO (31%), TRADE_NAME_DBA (9.2%), PW_OTHER_SOURCE (9.0%), ORIGINAL_CERT_DATE (6.6%) and PW_TRACKING_NUMBER (0.1%). The share of rows with a value also differs between workbooks: WORKSITE_COUNTY (100% in FY2024 Q4, 17% in FY2025 Q4 and 100% in FY2026 Q3).
DOL names each workbook for a federal fiscal year, which runs from October 1 to September 30, and a quarter. Fiscal 2025 began on October 1, 2024. DOL's performance data page lists four workbooks for each of fiscal 2024 and 2025, Q1 to Q4. The workbooks in our copy:
The record layout for fiscal 2025 gives the reporting period as October 1, 2024 through September 30, 2025. The FY2025 Q4 workbook holds the decisions of July 1, 2025 to September 30, 2025. Read the first and last DECISION_DATE of a workbook before counting a year from it. To cover a full fiscal year, collect the workbooks of the other quarters, read the dates of each and keep one row per case number.
An application decided in a window can have arrived long before: the earliest RECEIVED_DATE in the FY2024 Q4 workbook is November 18, 2019. Count a period by DECISION_DATE. DOL adds that a small percentage of determinations can change in later releases after an appeal or redetermination. The pages of this site describe 15 months of DOL decisions (Jul–Sep 2024, Jul–Sep 2025, Oct 2025–Jun 2026).
CASE_STATUS holds the status at the last significant event or decision. The workbooks in our copy hold 676,973 applications in four statuses:
ORIGINAL_CERT_DATE is filled on the 43,698 Certified - Withdrawn rows and on no others. A withdrawal comes a median of 337.5 days after certification: 16% within 30 days and 65% after more than 180.
The wage and sponsorship pages of this site count the first two statuses, 661,885 applications, since each records a wage that DOL certified at some point. A denial is rare: the seven-day review looks for completeness and obvious errors, and a certification records that an application passed it.
The offered wage is WAGE_RATE_OF_PAY_FROM in the unit of WAGE_UNIT_OF_PAY. Among the certified applications the units are Year 613,403, Hour 46,756, Month 1,127, Week 342 and Bi-Weekly 257, so 93% state a yearly wage and 7.1% an hourly one. Convert every wage to one unit before comparing. This site multiplies an hourly wage by 2,080 hours.
WAGE_RATE_OF_PAY_TO holds the top of a range. It is blank on 69% of the certified applications and sits above the lower figure on 31%. This site uses the lower figure as the offered wage.
DOL notes that typographical errors and other anomalies in the files can come from data entry, and some wages show it. The mean offered wage in our copy is $309,013 against a median of $125,000, because 1,138 applications annualize to $1,000,000 or more. Without them the mean is $136,025. The largest, for cardiologists, financial managers and chief executives, annualize to $1,000,731,680, $936,000,000 and $832,000,000. Divided by 2,080 hours, they give hourly figures of $400,000 to $481,121, amounts that read as annual salaries typed into an hourly field. A median or a percentile stays near the middle of the data. A mean follows these rows.
An application names its prevailing wage in one of three ways, and the columns show which. 91% of the certified applications carry an OES wage level in PW_WAGE_LEVEL, 864 carry a DOL tracking number in PW_TRACKING_NUMBER and 59,473 (9.0%) name another source in PW_OTHER_SOURCE: Survey 50,331, CBA 9,067, SCA 48 and DBA 27. Each certified application fills exactly one of the three.
PW_WAGE_LEVEL holds the level the employer placed the job in: I, II, III, IV or N/A. The mix in our copy:
Certified applications by wage level, share of 661,885
A blank level goes with another route to the prevailing wage: the 60,337 blank rows take their wage from a DOL determination or another source. The level is the employer's own placement of the job, and the H-1B selection now weights registrations by wage level, with one entry at level I and four at level IV. What each level pays shows the offered wages behind the levels.
Employers file under their legal names, and one company can file under several. Our copy holds 79,663 distinct employer names, and grouping the names of one company gives 71,614 employers. The ten largest hold 15% of the certified applications. The five largest: Amazon.com Services LLC (29,374), Cognizant Technology Solutions US Corp (10,957), Microsoft Corporation (9,885), Google LLC (9,809) and Apple Inc. (8,419).
The tax ID is the cleaner key. EMPLOYER_FEIN is filled on 100% of the certified applications, and the 79,663 names that carry one sit on 69,464 distinct tax IDs. 8,368 tax IDs appear under more than one name, and 350 names appear with more than one tax ID. A subsidiary has a tax ID of its own, so a tax ID separates a parent from its subsidiaries where a brand name joins them.
The employer's address and the worksite can differ. Of the certified applications that give both states, 47% list a worksite state other than the employer's. 19% say the workers will be placed with a secondary entity, 23% come from employers that report being H-1B dependent and 74% name an agent or attorney who represents the employer.
60% of the certified applications are in computer occupations, SOC codes that start with 15-. The five largest occupations: Software Developers (199,457), Computer Systems Engineers/Architects (54,561), Data Scientists (39,386), Software Quality Assurance Analysts and Testers (21,475) and Computer Programmers (16,891). Worksites reach 55 states and territories.
DOL's data page describes each data set as holding unique records identified by case number, based on the most recent date a determination was issued. A case certified in one period and withdrawn in a later one then moves to the later workbook. A case number appears once in a workbook. 3,735 case numbers in our copy appear in two of the workbooks. In every one of them the earlier workbook holds the row as Certified and the later workbook as Certified - Withdrawn. In every one of the 3,735 pairs, the later row's ORIGINAL_CERT_DATE is the earlier row's DECISION_DATE.
3,733 of the 3,735 name the same employer in both rows. Our 661,885 rows hold 658,150 distinct case numbers, so this site counts each of these cases in both workbooks it appears in. A reader who wants each certification once keeps one row per case number; the later row carries the certification day in ORIGINAL_CERT_DATE.
The H-1B sponsorship page and each employer page on this site show these filings by employer.
A certified LCA lets the employer go on to the petition. USCIS decides the petition, and the visa follows from there.
An LCA can request several workers: 27,627 of the 661,885 certified applications in our copy request more than one. 30% of the requested positions are for new employment.
An LCA file counts applications and requested positions. USCIS counts petitions it decided. This site counts the certified applications of 15 months of DOL decisions (Jul–Sep 2024, Jul–Sep 2025, Oct 2025–Jun 2026), with a company's names grouped.
Group by EMPLOYER_FEIN first. It is filled on 100% of the certified applications, and the legal name on its own can change from one filing to the next.
From the employer. PW_WAGE_LEVEL is the level the employer placed the job in when it set the OES wage itself, and a blank level goes with a DOL determination or another wage source.
H-1B LCA disclosure data: the quarterly workbooks of DOL's Office of Foreign Labor Certification (FY2024 Q4, FY2025 Q4 and FY2026 Q3), downloaded September 19, 2026
and read in full for this guide. Certified means a CASE_STATUS of Certified or Certified - Withdrawn. Counts of certified applications are those of the wage warehouse built September 29, 2026 from the same
workbooks; shares of certified applications are taken over the certified rows of the workbooks, and the share of case numbers over all applications. An employer is a company's names grouped by name, with a hand-approved family list.
Tax ID counts use the certified rows that carry a tax ID. An annualized wage multiplies an hourly wage by 2,080 hours. DOL's wording was read on October 19, 2026.
Built by scripts/build_insights.py.
Sources: DOL OFLC performance data; LCA record layout, fiscal year 2025; DOL H-1B, H-1B1 and E-3 programs; USCIS H-1B Employer Data Hub.
Informational only; not legal, immigration, or financial advice.
Look up any employer's H-1B filings, layoff notices and wage-and-hour record, with your own target salary range.
H-1B sponsorship by employer → · Where our other numbers come from → · Check a job or an employer →Get new articles and the biggest layoff notices
One email a week, sent only when there is something new.
Every message has an unsubscribe link. How we use your address
Amazon: 9,746 workers in 30 layoff notices
What state layoff notices and Labor Department H-1B filings show for Amazon: the notices over the past year, the sites, and the sponsorship record.
How to read BEA regional price parities: 387 metro areas
What a regional price parity of 95 or 110 means, how BEA builds one, how 387 metro areas spread around 100, why rent drives it and how the site uses it for pay after cost of living.
How to read OEWS wage data: 824 occupations in 393 metros
The columns and symbols of the OEWS workbooks, what a wage and a percentile mean, how precise the numbers are, and how the H-1B wage levels are built from them.
How to read DOL wage and hour enforcement data: 267,635 cases
The fields of a Wage and Hour case, the laws behind the numbers, what back wages agreed to pay and penalties assessed mean, and the cases a copy of the dataset leaves out.
How to read a WARN notice: what 4,099 state filings contain
What the notices can say, why some are missing, the federal rules and the state lists, the fields of a WARN notice and how we count them, with what a year of notices shows.
Kroger: 19 layoff notices naming 3,066 workers
The state layoff notices filed by Kroger, its warehouse company and its grocery chains: 19 notices and at least 3,066 workers in a year, with the wage and H-1B records alongside.
Wage data: U.S. Department of Labor OFLC LCA Disclosure Data and U.S. Bureau of Labor Statistics OEWS. LCA filings reflect sponsored-worker wage offers, which skew toward mid-large employers and specific occupations — treat any estimate built from a small sample cautiously. Informational only; not legal, immigration, or financial advice.