Every H-1B, H-1B1, and E-3 visa petition starts with a Labor Condition Application (LCA), in which the sponsoring employer tells the U.S. Department of Labor where the foreign worker will actually be employed. Federal regulation requires that the worksite listed on Form ETA-9035 be the physical location where the work is performed; it cannot be a P.O. Box.
To see how often that rule is followed, we compared more than 437,000 LCA records from the Department of Labor’s FY2026 Q3 public-disclosure data with addresses for virtual mailboxes and P.O. Boxes supplied by PostScan Mail, iPostal1, and Anytime Mailbox in California, Texas, and New York, the three states that file the most H-1B petitions. The comparison turned up dozens of employers listing a commercial mailbox as the worksite for a sponsored worker, a practice that could violate federal law and resembles patterns seen in past H-1B visa-violation prosecutions.
Background
What is the LCA (Labor Condition Application)?
Before an employer can bring a foreign professional to the U.S. on an H-1B, H-1B1, or E-3 visa, it must file a Labor Condition Application (LCA) with the Department of Labor. The LCA is meant to keep the hiring of a foreign worker from uncompetitive wages or working conditions for U.S. workers. It covers three visa categories: the H-1B (up to 65,000 visas annually, plus an additional 20,000 under the advanced-degree exemption), the H-1B1 (6,800 visas), and the E-3 (10,500 visas).
The LCA process has three steps. First, the employer determines the prevailing wage for the position and agrees to pay either that amount or the actual wage paid to similarly employed workers at the company, whichever is higher. Second, the employer submits Form ETA-9035/9035E electronically through the Department of Labor’s FLAG (Foreign Labor Application Gateway) system no more than six months before the worker’s start date. Third, the Department reviews the application for completeness, typically within seven business days; it does not assess wage or worksite claims. Once approved, the employer can file a visa petition with USCIS.
Worksite address requirement
The LCA does not simply ask an employer to name a city or state; it requires a precise, verifiable street address, and federal regulation leaves no ambiguity about what that address must be. The instructions accompanying Form ETA-9035 and its companion guidance, Form ETA-9035CP, both published by the Department of Labor, state:
“Each intended place of employment listed below must be the worksite or physical location where the work will actually be performed and cannot be a P.O. Box.”
That’s a direct requirement grounded in federal regulation, not guidance or best practice. Under 20 CFR § 655.715, a “place of employment” is legally defined as the worksite or physical location where the H-1B, H-1B1, or E-3 worker actually performs their job. 20 CFR § 655.730(c)(4)–(5) further requires the employer to identify every intended worksite on the LCA, including short-term placements, with as much geographic specificity as possible.
When an employer files an LCA, it is making a sworn representation to the federal government that a named commercial address is where a real person will physically show up to work. A P.O. Box, a mail-forwarding suite, or a virtual office with no actual workspace does not satisfy that requirement, regardless of whether the employer lists a suite number, omits one, or lists one that doesn’t match the mailbox provider’s own numbering.
That distinction between an address that merely receives mail and one where work is actually performed is the foundation of this investigation.
Penalties for Worksite Misrepresentation on an LCA
Listing a P.O. Box, a virtual mailbox, or any other fictitious address as an H-1B worker’s worksite isn’t a paperwork technicality. It’s a misrepresentation of a material fact, and it carries real exposure on two separate fronts: civil enforcement by the Department of Labor, and, where the false statement is part of a broader scheme, federal criminal prosecution.
Civil penalties (DOL Wage and Hour Division). Under 20 CFR §§ 655.805 and 655.810 and INA § 212(n)(2)(C), the DOL’s Wage and Hour Division can assess civil money penalties on a tiered scale, adjusted annually for inflation:
- Tier 1 ($2,364 per violation): a substantial violation of the LCA’s specificity requirements, including misrepresentation of a material fact such as the worksite.
- Tier 2 ($9,624 per violation): a willful violation or willful misrepresentation.
- Tier 3 ($67,367 per violation): willful violation combined with the displacement of a U.S. worker.
These figures are current under 20 CFR § 655.810(b) as adjusted effective January 10, 2025 (90 FR 1858); the Department of Labor anno unced no further inflation adjustment for 2026, so the amounts above still apply.
Beyond fines, a confirmed violation triggers mandatory debarment: a 1-year minimum bar from filing new H-1B, H-1B1, or E-3 petitions, rising to 2 years for a willful violation and 3 years where U.S.-worker displacement is involved. The employer also remains on the hook for back wages, the gap between what the worker should have been paid and what they actually received, independent of any penalty.
Criminal exposure. Where a false worksite address is part of a deliberate scheme, for example, to bill third-party clients while telling DOL the worker would be employed in-house, it can support federal charges of visa fraud (18 U.S.C. § 1546(a)), often paired with conspiracy (18 U.S.C. § 371) and wire fraud (18 U.S.C. § 1343). Courts have consistently held that a false statement about where and for whom an H-1B worker will actually work is legally “material” to the government’s visa decision, even when the employer argues immigration officials had no right to ask for that level of detail in the first place.
Court Precedents: How Judges Have Treated H-1B Worksite Violations
Federal courts have repeatedly confronted and punished this fact pattern: an employer telling the government one thing about where a foreign worker will be employed while doing another.
United States v. Nanda, 867 F.3d 522 (5th Cir. 2017). Brothers Atul and Jiten Nanda ran Dibon Solutions, a Texas IT staffing firm that told DOL and USCIS that roughly 200 H-1B workers would be employed directly at Dibon’s Carrollton headquarters. In reality, the company benched workers, paying them nothing or requiring them to cover their own expenses, until it could quietly place them with undisclosed third-party clients and pocket a cut of their billed rate. The Fifth Circuit affirmed convictions for conspiracy to commit visa fraud and wire fraud, sentencing each brother to 87 months in prison.
United States v. Patnaik, 125 F.4th 1223 (9th Cir. 2025). Executives of PerfectVIPs, Inc. were charged with filing 85 fraudulent H-1B applications that falsely claimed workers would be employed onsite on internal projects when they were actually contracted out to other companies. A district court initially dismissed the indictment, reasoning that under a 2020 ruling from the U.S. District Court for the District of Columbia (ITServe Alliance v. Cissna), USCIS could not demand that level of worksite detail in the first place, so the false statement could not be considered material. The Ninth Circuit reversed, holding that lying about where an H-1B worker will actually be employed remains criminal visa fraud regardless of whether the government was entitled to ask the question. That closes off a defense theory that shell or false worksite claims are legally harmless.
Cyberworld Enterprise Technologies, Inc. v. Napolitano, 602 F.3d 189 (3d Cir. 2010). An H-1B staffing company (doing business as Tekstrom) placed workers with client employers without making the legally required inquiry into whether doing so would displace a U.S. worker. DOL assessed a civil penalty and imposed a mandatory one-year debarment under INA § 212(n)(2)(C)(i)(II); the Third Circuit affirmed, confirming that once a violation is found, debarment follows automatically, with no proof of the employer’s bad intent required.
Read together, these cases show that courts have consistently punished false statements about where and for whom an H-1B worker actually works, but that is not the same as saying any of them is a P.O. Box case on its facts. Whether a specific virtual-mailbox worksite listing in this report would itself support a prosecution or civil penalty depends on additional facts, such as intent, pattern, and financial harm, that only a full investigation, not an address-matching exercise, can establish.
Data Sources Used in This Report
LCA Data
We examined the FY2026 Q3 disclosure file, which covers LCAs approved between October 1, 2025, and June 30, 2026. After deleting a set of blank rows that Excel added to the spreadsheet’s used range, the file contains 437,496 valid LCA records.
Each record is a single labor certification that holds about 100 fields, described in the layout. For this report only a handful are relevant:
- Sponsor – the employer’s name and business address, plus the attorney or agent who filed the LCA.
- Job details – title, occupation code, and whether the position is full‑time or part‑time.
- Visa type – H‑1B, H‑1B1, or E‑3, together with certification status (certified, denied, withdrawn).
- Worksite address – street address, city, county, state, and ZIP code where the foreign worker is expected to report.
The worksite field is central to this analysis. DOL also publishes a column‑definition file that we used to confirm exactly how each worksite attribute is defined before running any queries.
Addresses Data
To test whether any LCA worksite addresses trace back to a mailbox rather than a real office, we needed a reference list of known virtual-address and P.O. Box-style locations to check against. We built that list from three of the largest commercial mail-forwarding and virtual-office providers operating in the United States:
All three companies openly advertise their available locations to prospective customers. Anyone can browse their websites and see exactly which street addresses they offer, city by city and state by state. Nothing in this reference list was obtained through a non-public database, a subscription service, or any authenticated source; it reflects only what these providers themselves choose to publish as available mailbox and virtual-office locations for rent.
We compiled these public listings into a single combined address list for each state, removing duplicate entries where the same physical address appeared more than once, whether listed by two different providers or repeated within one provider’s own listings. The result is a clean, state-by-state inventory of real-world addresses that function as commercial mail drops rather than workplaces, which became the basis for cross-referencing against the LCA worksite data.
You can download all addresses we collected and processed from virtual office providers here in a CSV file:
Methodology
Matching worksite addresses
Comparing two large address lists sounds simple in theory: check whether an address in one list also appears in the other. But real-world addresses are written inconsistently enough that a literal, word-for-word comparison would miss the overwhelming majority of genuine matches.
The same physical address can appear as “St” in one filing and “Street” in another, “Blvd” versus “Boulevard,” “N” versus “North.” Before comparing anything, we standardized both the LCA worksite addresses and the mailbox-provider addresses into a consistent format: expanding abbreviations, stripping punctuation, and correcting for the different ways an address, city, state, and ZIP code can be strung together across three separate data sources.
We also made a deliberate judgment call on suite numbers. If an LCA lists “123 Main St” with no suite number, and a mailbox provider operates “123 Main St, Suite 200,” we still count that as a match. Employers frequently omit suite numbers on the LCA even when a real one exists, and a virtual mailbox provider can receive mail addressed to their building regardless of which specific suite is named. Requiring an exact suite match would have let genuine matches slip through simply because of incomplete paperwork, so we matched on the street address and ZIP code and ignored the suite number entirely.
Finally, we compared each mailbox-provider address only against the LCA’s declared worksite, where the law requires the work to actually happen, not against the employer’s own business mailing address. A company’s headquarters being a virtual office is a common and largely unremarkable business practice. Whether the worker is being told to report to a mail drop instead of a real workplace is a different question.
Filtering out shared office buildings and false positives
The single biggest source of misleading matches turned out to be large, well-known office towers. A handful of these buildings, the kind that house dozens of unrelated tenants, from Fortune 500 companies to small law firms, happen to also contain a virtual-mailbox provider’s suite somewhere inside. When we found companies like Amazon and Accenture “matching” a mailbox address, it wasn’t evidence of a violation. It was evidence that a single office tower can hold both a legitimate multinational’s real office and, several floors away, a completely unrelated mail-forwarding business.
A simple rule of excluding any address shared by more than one employer cuts both ways, though: it would also hide the opposite pattern, several small, unrelated-looking shell companies clustered at one mailbox, which is arguably more suspicious than a single filer, not less. So instead of a flat headcount, we applied a data-driven test. For every address shared by multiple employers, we checked each of those employers against the entire nationwide LCA dataset, not just the matched subset, to see whether they file from any other worksite address anywhere in the country. If every employer at a shared address is confined to that one address nationwide, the address stays flagged, on the theory that a cluster of single-location companies at one mailbox is at least as suspicious as a lone filer. If any employer at that address also operates at other, different worksites elsewhere, the address is set aside, since that employer’s footprint at multiple real locations is evidence the building itself is an ordinary multi-tenant office, not a violation pattern.
Limitations and data integrity checks
This process is a screening tool, not a courtroom-ready determination of fraud or any other violations. A matched address means an employer told the federal government, under a legally binding attestation, that a foreign worker would be performing real work at a location that is, in fact, a commercial mailbox. That is a flaggable discrepancy, but a match might also reflect a data-entry error rather than deliberate misrepresentation. Every result in this report should be read as a candidate for further human and regulatory review, not a final legal conclusion.
What We Found
California
After the shared-building filter, 68 LCA records across 54 distinct employers in 30 different California cities were flagged with a worksite address matching a known virtual-mailbox location. iPostal1 accounted for the largest share of matched addresses (48), followed by Anytime Mailbox (27) and PostScan Mail (9) — some addresses appear in more than one provider’s listings.
The flagged employers span a wide range of business types — engineering consultancies, healthcare staffing, and numerous small software and IT-consulting companies — rather than clustering in one obvious industry. Sumedha Global Computing, Inc., for example, filed five separate H-1B applications at “901 H Street, Ste. 130, Sacramento, CA 95814,” an address matching an iPostal1 listing. Appleton Luff Trade Consulting LLC filed three at “1440 Broadway, Suite 208-G, Oakland, CA 94612,” also matching an iPostal1 listing.
California — Flagged H-1B Worksite Matches
The table below lists 68 certified, withdrawn, or denied Labor Condition Applications whose declared California worksite address matches a known virtual-mailbox or P.O. Box location from PostScan Mail, iPostal1, or Anytime Mailbox — after excluding addresses shared by more than one unrelated employer (see our methodology for how and why). Every column from the original Department of Labor disclosure file is shown below; use the search box to filter across all of them, or scroll horizontally to see more columns.
Note: A matched address means the employer told the U.S. Department of Labor, under a legally binding attestation, that a foreign worker would be performing real work at a location that is, in fact, a commercial mailbox — a discrepancy worth flagging, not a legal finding of fraud. Some matches may reflect a data-entry error rather than intentional misrepresentation. Verify any record against the Department of Labor’s public disclosure data before drawing conclusions about any specific employer.
Texas
Texas produced 86 flagged records across 26 distinct employers in 10 cities — after excluding addresses where a genuinely distinct, multi-location employer was also present, including a well-known Dallas office tower shared by Amazon, Accenture, and several other tenants with real operations at other locations nationwide (see Methodology).
The single strongest pattern in this investigation is AR Infotek Solutions LLC, a Texas company that filed 52 separate H-1B applications — 51 of them certified — for jobs ranging from Data Analyst to Network Engineer to Power BI Developer, every one of them listing the identical worksite: “501 S. Austin Ave, #1220, Georgetown, TX 78626.” That exact street address and suite number matches an iPostal1 virtual-mailbox listing precisely — not an approximate, suite-optional match, but an exact one. Fifty-one certified H-1B workers, on paper, all report to the same rented mailbox suite.
Two smaller but similar patterns: STELMITH LLC filed three separate LCAs at “13612 Midway Road, Ste-118, Dallas, TX 75244,” and Digital Janet LLC filed two at “17250 Dallas Pkwy, Dallas, TX 75248” — both addresses matching iPostal1 listings. Filing many LCAs from a single small company at one virtual-mailbox address is a stronger version of the pattern this investigation was designed to surface, compared to a one-off filing.
Texas — Flagged H-1B Worksite Matches
The table below lists 86 certified, withdrawn, or denied Labor Condition Applications whose declared Texas worksite address matches a known virtual-mailbox or P.O. Box location from PostScan Mail, iPostal1, or Anytime Mailbox — after excluding addresses shared by more than one unrelated employer (see our methodology for how and why). Every column from the original Department of Labor disclosure file is shown below; use the search box to filter across all of them, or scroll horizontally to see more columns.
Note: A matched address means the employer told the U.S. Department of Labor, under a legally binding attestation, that a foreign worker would be performing real work at a location that is, in fact, a commercial mailbox — a discrepancy worth flagging, not a legal finding of fraud. Some matches may reflect a data-entry error rather than intentional misrepresentation. Verify any record against the Department of Labor’s public disclosure data before drawing conclusions about any specific employer.
New York
New York produced 30 flagged records across 21 employers in 6 cities. iPostal1 again dominated the matched addresses (26 of 30).
QUANTIFI, INC. filed four separate H-1B applications at “276 5th Avenue, Suite 1103, New York, NY 10001” — an address matching an iPostal1 listing — the most repeat filings from a single employer found in New York.
LBDO GROUP INC similarly appears twice, filing both a Director of Business Development and a Business Development and Partnerships role at “1 Carlton Avenue, #1009, Brooklyn, NY 11205.” Notably, one flagged record, a Child Psychotherapist position at a Scarsdale address, carries a case status of “Withdrawn” rather than “Certified” — a reminder that not every flagged filing resulted in an active visa; withdrawn and denied cases remain in the dataset and in this report for completeness, but they did not result in DOL certifying the worksite claim.
New York — Flagged H-1B Worksite Matches
The table below lists 30 certified, withdrawn, or denied Labor Condition Applications whose declared New York worksite address matches a known virtual-mailbox or P.O. Box location from PostScan Mail, iPostal1, or Anytime Mailbox — after excluding addresses shared by more than one unrelated employer (see our methodology for how and why). Every column from the original Department of Labor disclosure file is shown below; use the search box to filter across all of them, or scroll horizontally to see more columns.
Note: A matched address means the employer told the U.S. Department of Labor, under a legally binding attestation, that a foreign worker would be performing real work at a location that is, in fact, a commercial mailbox — a discrepancy worth flagging, not a legal finding of fraud. Some matches may reflect a data-entry error rather than intentional misrepresentation. Verify any record against the Department of Labor’s public disclosure data before drawing conclusions about any specific employer.
Conclusion
Federal law is unambiguous: the worksite listed on an H-1B, H-1B1, or E-3 Labor Condition Application must be the real, physical location where a foreign worker will perform their job, not a mailbox. Cross-referencing over 437,000 LCA records (certified, denied, and withdrawn alike) against three major virtual-mailbox providers surfaced 184 flagged records across California, Texas, and New York, after excluding addresses where a genuinely distinct, multi-location employer was also present (see Methodology). These are not hypothetical addresses; they are specific, publicly filed government records tied to specific companies, specific job titles, and specific case numbers.
A company repeatedly listing a commercial mail-forwarding address as the “workplace” for dozens of sponsored workers is precisely the fact pattern that warrants further public scrutiny and possible investigation. This report’s methodology cannot by itself establish that any individual filing was fraudulent rather than mistaken. Whether any individual flagged filing here reflects an innocent data-entry choice, a legitimate short-term or remote arrangement, or an actual misrepresentation is not something a data-matching exercise can determine on its own. What this report does establish is that the problem is real, it is presently detectable using only public information, and it warrants closer review, by journalists, by the employers named, and by the federal agencies with actual investigative and subpoena authority.
Limitations and disclaimers
Beyond the methodology caveats already discussed (that this is a screening tool rather than a legal finding, that suite-number mismatches were treated as non-disqualifying, and that shared-building matches were evaluated individually rather than assumed innocent), a few additional limits on this report’s scope should be kept in mind:
Partial fiscal year. The underlying data covers FY2026 Q3 only, October 2025 through June 2026, not a full fiscal year. A full-year dataset would likely surface additional matches simply due to a larger volume of filings.
Three states, three providers. This report examined only California, Texas, and New York (the top three states by LCA filing volume) and only three virtual-mailbox providers (PostScan Mail, iPostal1, Anytime Mailbox). Numerous other commercial mail-forwarding services, including UPS Store locations, Regus/IWG virtual offices, and smaller regional providers, were not included.
First worksite only. DOL’s disclosure file, and this analysis, captures only each LCA’s primary, first-listed worksite. Some employers report additional worksite locations in a separate DOL file not used in this report, meaning some address discrepancies may not have been captured here at all.
Self-reported, unverified data. DOL certifies an LCA for completeness, not for the truthfulness of the worksite claim at the time of filing. Every address in the underlying dataset was reported by the employer itself; DOL does not independently verify or geocode worksite addresses before certification, and neither did this report beyond the text-matching described in the Methodology.
Employer-name matching, not fuzzy company research. Two filings sharing an address were treated as the same company only where they share a federal EIN, or an exact (punctuation- and case-normalized) legal name. Genuinely affiliated businesses filing under different legal names and different EINs at the same address would be counted as separate companies by this methodology, and a coincidental name collision between two unrelated companies would not be caught by the name-based fallback.
Not an official investigation. This report was produced independently using publicly available data and is not affiliated with, endorsed by, or conducted on behalf of the Department of Labor, USCIS, or any law enforcement agency. It is intended to inform the public and prompt further review by those with the legal authority to investigate individual cases, not to serve as a final adjudication of wrongdoing by any employer named.
A note on the employers named in this report. Every fact stated about a named employer in this report, its name, address, job title, and case number, comes directly from that employer’s own Labor Condition Application, a public record the employer itself filed with the U.S. Department of Labor under penalty of perjury. Naming an employer here means only that a publicly listed address on that employer’s LCA filing matched a publicly listed address of a commercial virtual-mailbox or P.O. Box provider, using the automated methodology described above, nothing more. It is not an allegation, finding, accusation, or implication that any named employer has committed fraud, violated any law or regulation, or acted with improper intent, and no individualized investigation of any named employer’s actual business operations, physical presence, or state of mind was performed beyond that address comparison. DOL certification of an LCA reflects the Department’s completeness review at the time of filing; it is neither proof that the worksite claim was accurate nor evidence that it was not. A match identified here may have a fully lawful explanation this report’s automated methodology cannot detect, including a registered-agent or short-term filing arrangement, a data-entry error on the original filing, a since-corrected record, or a legitimate business use of the address in question. Any employer named in this report who believes their inclusion is inaccurate, outdated, or based on an error in the underlying public data is invited to contact us at info@visainamerica.com with supporting documentation; this report will be corrected promptly upon verification. This report is public-interest commentary on matters of public record concerning the integrity of a federal visa program, and nothing in it should be read to state or imply that any named employer has been charged with, is under investigation for, or has been found liable for any crime or civil violation.
