TL;DR. 24,360 US employers won 81,497 H-1B approvals in FY2026 — down 54% from FY2023's 176,949. Amazon alone drove 2,448 of them, mostly out of Arlington, VA. Microsoft drove 1,174 out of Redmond. Google drove 1,027 out of Mountain View. If you need sponsorship, these 25 cities are where the jobs are — and the gap between the "HQ-registered" ranking and where people actually sit at a desk matters more than almost anyone writing about H-1B tells you.
The dataset
Every row of USCIS employer-level H-1B approval data for FY2026, joined against DOL LCA worksite disclosures for the same fiscal year. We counted approvals (not filings) because an approval is a job that actually got sponsored. Wages come from the LCA table where both the annualized wage and the worksite city are populated.
• Employers with FY2026 approvals: 24,360
• FY2026 approvals total: 81,497
• FY2023 approvals total (same methodology): 176,949
• YoY change: -54%
• LCA worksite filings (FY2026, wage sanity-checked): 437,496
• National FY2026 approval rate: 97.1%
Methodology at the bottom.
Top 15 cities by H-1B approvals (FY2026)
Grouped by the employer's registered city (not worksite). The right-hand column tells you how brutally the YoY drop hit:
| City | FY2026 approvals | Sponsor companies | Approval rate | vs FY2023 |
|---|---|---|---|---|
| New York, NY | 4,834 | 1,502 | 97.0% | -42.7% |
| Arlington, VA | 2,840 | 73 | 98.0% | -53.2% |
| Chicago, IL | 2,829 | 462 | 97.6% | -53.0% |
| San Francisco, CA | 2,088 | 546 | 97.4% | -45.4% |
| San Jose, CA | 1,795 | 264 | 98.1% | -48.9% |
| Santa Clara, CA | 1,786 | 140 | 98.2% | -16.1% |
| Rockville, MD | 1,686 | 45 | 95.4% | -57.2% |
| Austin, TX | 1,610 | 246 | 96.3% | -29.9% |
| Mountain View, CA | 1,503 | 94 | 98.0% | -60.6% |
| Boston, MA | 1,489 | 300 | 97.0% | -32.2% |
| Richardson, TX | 1,363 | 72 | 99.3% | -64.5% |
| Redmond, WA | 1,282 | 36 | 96.8% | -39.2% |
| Philadelphia, PA | 1,038 | 101 | 98.4% | -61.4% |
| College Station, TX | 1,006 | 11 | 98.5% | -71.7% |
| Cupertino, CA | 1,006 | 23 | 98.5% | -44.5% |
New York, NY
- FY2026 approvals
- 4,834
- Sponsor companies
- 1,502
- Approval rate
- 97.0%
- vs FY2023
- -42.7%
Arlington, VA
- FY2026 approvals
- 2,840
- Sponsor companies
- 73
- Approval rate
- 98.0%
- vs FY2023
- -53.2%
Chicago, IL
- FY2026 approvals
- 2,829
- Sponsor companies
- 462
- Approval rate
- 97.6%
- vs FY2023
- -53.0%
San Francisco, CA
- FY2026 approvals
- 2,088
- Sponsor companies
- 546
- Approval rate
- 97.4%
- vs FY2023
- -45.4%
San Jose, CA
- FY2026 approvals
- 1,795
- Sponsor companies
- 264
- Approval rate
- 98.1%
- vs FY2023
- -48.9%
Santa Clara, CA
- FY2026 approvals
- 1,786
- Sponsor companies
- 140
- Approval rate
- 98.2%
- vs FY2023
- -16.1%
Rockville, MD
- FY2026 approvals
- 1,686
- Sponsor companies
- 45
- Approval rate
- 95.4%
- vs FY2023
- -57.2%
Austin, TX
- FY2026 approvals
- 1,610
- Sponsor companies
- 246
- Approval rate
- 96.3%
- vs FY2023
- -29.9%
Mountain View, CA
- FY2026 approvals
- 1,503
- Sponsor companies
- 94
- Approval rate
- 98.0%
- vs FY2023
- -60.6%
Boston, MA
- FY2026 approvals
- 1,489
- Sponsor companies
- 300
- Approval rate
- 97.0%
- vs FY2023
- -32.2%
Richardson, TX
- FY2026 approvals
- 1,363
- Sponsor companies
- 72
- Approval rate
- 99.3%
- vs FY2023
- -64.5%
Redmond, WA
- FY2026 approvals
- 1,282
- Sponsor companies
- 36
- Approval rate
- 96.8%
- vs FY2023
- -39.2%
Philadelphia, PA
- FY2026 approvals
- 1,038
- Sponsor companies
- 101
- Approval rate
- 98.4%
- vs FY2023
- -61.4%
College Station, TX
- FY2026 approvals
- 1,006
- Sponsor companies
- 11
- Approval rate
- 98.5%
- vs FY2023
- -71.7%
Cupertino, CA
- FY2026 approvals
- 1,006
- Sponsor companies
- 23
- Approval rate
- 98.5%
- vs FY2023
- -44.5%
Every top-15 city contracted. The steepest falls — College Station, Richardson, Philadelphia, Rockville — all share the same shape: one or two Indian IT-services giants (Cognizant, Infosys, Deloitte, TCS) pulled their petition budget back hard. The cities where product companies run engineering (Santa Clara, Austin, Boston) held up the best.
There are two kinds of H-1B cities
Look at the "sponsor companies" column in the table above. The spread is wild — New York runs 1,502 sponsors, College Station runs 11. That is not a long-tail artifact. It is the single most important fact about the H-1B map, and it splits every city into one of two archetypes.
1. Captive hubs — one employer is the city
These cities have fewer than 150 active sponsors, and one of them accounts for the overwhelming majority of approvals. If that company freezes its cap-season hiring, the whole city's number craters.
Arlington, VA — Amazon (1,958 of 2,840 approvals = 69%). HQ2.
Rockville, MD — Tata Consultancy Services (1,490 of 1,686 = 88%).
Mountain View, CA — Google (1,027 of 1,503 = 68%).
Richardson, TX — Infosys (1,136 of 1,363 = 83%).
Redmond, WA — Microsoft (1,174 of 1,282 = 92%).
College Station, TX — Cognizant (867 of 1,006 = 86%).
Cupertino, CA — Apple (982 of 1,006 = 98%).
Applying to a job in Cupertino that is not at Apple? There are roughly 24 approved H-1B sponsors in the whole city. Plan accordingly.
2. Diversified metros — hundreds of sponsors, no single king
These cities spread approvals across many employers. No single company is above 15% of the city total. These are the safe default for an H-1B job search because one company's hiring freeze does not break the whole market for you.
New York, NY — 1,502 sponsors. Top 3: Citibank (308), McKinsey (207), Barclays (166).
Chicago, IL — 462 sponsors. Top 3: JPMorgan Chase (614), Capgemini (455), Accenture (454).
San Francisco, CA — 546 sponsors. Top 3: Salesforce (288), Uber (121), DoorDash (98).
San Jose, CA — 264 sponsors. Top 3: Cisco (319), PayPal (219), Adobe (117).
Boston, MA — 300 sponsors. Top 3: Fidelity (398), BCG (69), State Street (55).
Austin, TX — 246 sponsors. Top 3: Tesla (625), Oracle (434), Samsung (39).
Where H-1B workers actually sit is not where their employer files
The USCIS table above is sorted by the employer's HQ address. That under-counts the cities where big-tech H-1B workforces physically report to work. When you use the DOL Labor Condition Application data instead — which records the actual worksite address on each sponsored job — the ranking shifts materially.
Top 15 cities by FY2026 LCA worksite filings, with median annualized wage:
| Worksite city | LCA filings | Median wage |
|---|---|---|
| New York, NY | 28,378 | $165,000 |
| Seattle, WA | 12,849 | $162,600 |
| Austin, TX | 11,171 | $140,000 |
| San Francisco, CA | 10,352 | $195,000 |
| San Jose, CA | 9,348 | $188,000 |
| Sunnyvale, CA | 8,353 | $196,000 |
| Chicago, IL | 7,423 | $122,477 |
| Bellevue, WA | 6,770 | $167,024 |
| Santa Clara, CA | 6,450 | $196,938 |
| Plano, TX | 6,389 | $115,000 |
| Dallas, TX | 6,389 | $116,200 |
| Houston, TX | 6,387 | $108,243 |
| Atlanta, GA | 6,356 | $120,000 |
| Irving, TX | 5,726 | $114,379 |
| Charlotte, NC | 5,392 | $120,111 |
New York, NY
- LCA filings
- 28,378
- Median wage
- $165,000
Seattle, WA
- LCA filings
- 12,849
- Median wage
- $162,600
Austin, TX
- LCA filings
- 11,171
- Median wage
- $140,000
San Francisco, CA
- LCA filings
- 10,352
- Median wage
- $195,000
San Jose, CA
- LCA filings
- 9,348
- Median wage
- $188,000
Sunnyvale, CA
- LCA filings
- 8,353
- Median wage
- $196,000
Chicago, IL
- LCA filings
- 7,423
- Median wage
- $122,477
Bellevue, WA
- LCA filings
- 6,770
- Median wage
- $167,024
Santa Clara, CA
- LCA filings
- 6,450
- Median wage
- $196,938
Plano, TX
- LCA filings
- 6,389
- Median wage
- $115,000
Dallas, TX
- LCA filings
- 6,389
- Median wage
- $116,200
Houston, TX
- LCA filings
- 6,387
- Median wage
- $108,243
Atlanta, GA
- LCA filings
- 6,356
- Median wage
- $120,000
Irving, TX
- LCA filings
- 5,726
- Median wage
- $114,379
Charlotte, NC
- LCA filings
- 5,392
- Median wage
- $120,111
Seattle appears at #2 with 12,849 filings but is not in the HQ-ranked table at all — Amazon registered 1,958 approvals under Arlington, VA, and those workers mostly live in Seattle. Sunnyvale and Bellevue tell the same story for the rest of Big Tech. Charlotte breaks into the top 15 because every national bank (BofA, Wells, Truist, Ally) stages H-1B hires there.
Implication for a job search: if you only target cities from the USCIS-HQ list, you miss the real worksite cluster for Amazon, Microsoft, and Meta. Target the LCA worksite list instead — or at least both.
The wage gap by city is enormous
The median annualized H-1B wage in the Bay Area trio (Santa Clara / Sunnyvale / San Francisco) is $188K–$197K. In the Texas trio (Plano / Dallas / Houston) it is $108K–$116K. Same visa, same fiscal year, same approval machinery — almost double the pay depending on where you are.
The Bay Area cluster is heavy with FAANG product teams (NVIDIA, Google, Apple, Meta, Palantir, Databricks). The Texas cluster is heavy with large-cap IT services and enterprise back-offices. Approval mechanics are identical; the jobs under them are different jobs.
If you are optimizing for wage, anchor your search in the LCA worksite-ranked top 5: New York, Seattle, Austin, San Francisco, San Jose. Everywhere else, you are taking a 30-50% median pay cut on the same visa.
Approval rates are almost identically high everywhere
Every top city clears 96–99% approval. The FY2026 national rate is 97.1%. The spread between the strictest-looking city (Rockville, 95.4%) and the loosest (Richardson, 99.3%) is only 3.9 points.
What that means in plain English: the USCIS adjudicator is not the bottleneck. The bottleneck is whether a given employer files for you in the first place. The 54% YoY contraction in approvals is demand-side — companies filed fewer petitions. Approval rates on the petitions they did file barely moved.
Which brings us to the single most important thing you can do in an H-1B job search.
The actual H-1B job-search move
Every section of this post rhymes with the same conclusion: whether an H-1B sponsor hires you is not a function of your resume quality. It is a function of whether you ever applied to a company that files petitions in the first place. Roughly 60-70% of US employers have never filed. If your applications are not filtered against the USCIS employer list, most of them are going to companies that will reject you on the visa question no matter how well the interview goes.
This is exactly the filtering layer we built HiringFunnel for. When a client tells us they need sponsorship, we sort every open US role by the parent employer's recency-weighted H-1B petition volume and approval rate — the same data behind the tables in this post — and only apply you to the top of that list. No applications go to companies that do not sponsor.
For the deeper walk-through of how the filtering works, see What If I Need H-1B Sponsorship? Here's How We Solve It and the tactical companion piece H-1B Software Engineer Job Search: How to Find Sponsors That Actually Hire.
About the research
Data source. USCIS H-1B Employer Data Hub for FY2026 (approvals and denials by employer, state, city, NAICS). DOL Office of Foreign Labor Certification LCA disclosure data for FY2026 (worksite city, employer city, wage range, job title). Both are public federal datasets.
What we counted. Approvals, not filings. An approval is a petition where USCIS actually granted sponsorship. A filing can be withdrawn, denied, or rejected on a technicality — counting filings inflates the picture. Where YoY comparisons are shown, both years run the same approval-only methodology.
Which city? For the USCIS table, employer-registered city. For the LCA worksite table, the city of the actual work location on the labor condition application. These two cities often differ for Fortune 500 employers with a single corporate address and nationwide engineering teams.
Wage. Annualized wage from the LCA record, filtered to the $30,000–$1,000,000 range to drop data-entry errors. Where a single LCA covers multiple positions, each position counts once at the recorded wage.
What is not in the set. Approvals filed outside FY2026 are only present as the FY2023 comparison point. FY2024 and FY2025 are not in our H-1B employer table. Approvals that USCIS did not attribute to a specific city are excluded. H-4 EAD and other dependent visas are out of scope.
Want the raw query, a different cut, or the sponsor list for a specific city? Email me.
If you need H-1B sponsorship and you want this kind of filtering run on your own search — only companies that sponsor, only recent filers, only roles that fit your resume — the application is open.