← FAT Research Library

📅 Prepared September 10, 2026   ✍️ Dirk Adams with the assistance of AI   ⌛ 14 min read   14 sources

FOREIGN MEAT IMPORTATION RESEARCH SERIES | RESEARCH PAPER

The Border Check

What USDA actually inspects on 6.5 billion pounds of imported meat, how it decides a foreign system is safe, and why almost none of it reaches your label.

Prepared for publication by Farm Animal Transparency | September 2026

The federal government requires imported meat to declare its country of origin at the dock, records it, and then does not require anyone to show it to the person who eats the product.

Abstract

This paper examines what the U.S. Department of Agriculture does with imported meat and poultry between the port and the point of sale, over the five years from 2021 to 2026. Using USDA’s own budget justifications, FSIS directives, and two USDA Office of Inspector General audits, it separates three things that are routinely conflated: the checks performed on every lot, the deeper examinations performed on a sample, and the determination that a foreign country’s inspection system is equivalent in the first place. It finds a system that inspects everything at a shallow depth, samples at greater depth, and rests its country-level judgments substantially on self-reported annual data between audits scheduled every two to three years — a defensible architecture that no independent auditor has stress-tested since conditions in FY2017–2019. It then locates the failure that matters most to a shopper, which is not at the border at all: origin is required on the imported carton and not required on the retail package, because mandatory country-of-origin labeling for beef and pork was repealed in 2015. This is a companion to Purchased Silence, which traces by name which American companies receive imported beef; this paper traces what the government does with it.

1. The question at the meat case

Stand in front of the beef case and ask where the animal lived. For chicken, lamb, goat, and fish, the law still requires an answer. For beef and pork — the two products Americans buy most — it does not. Congress repealed mandatory country-of-origin labeling for beef and pork in the Consolidated Appropriations Act, 2016, signed in December 2015, following a World Trade Organization ruling against the program. The requirement did not weaken. It ended, and chicken, lamb, goat, fish and shellfish, and a list of produce and nuts carried on without it.

The volume behind that silence keeps growing. The United States imported 5.471 billion pounds of beef in 2025 (carcass-weight equivalent), an increase of 835 million pounds — about 18% — over 2024. Australia supplied 296 million pounds more than the prior year, up 27%; Brazil 241 million more, up 35%; Mexico 114 million more, up 19%. New Zealand fell 4% and Canada was flat. USDA’s Economic Research Service forecast 5.675 billion pounds for 2026, roughly 4% higher again.

So the useful question is not whether imported meat is inspected. It is what the government checks, what it learns, and how much of that survives the trip to the label.

2. What happens at the border

Imported meat is not waved through. Product offered for importation must bear the foreign establishment’s mark and the name of the country of origin preceded by the words “Product of,” and every shipment presented at an official import inspection establishment receives routine reinspection: certification and eligibility verification, label checks, container condition, evidence of shipping damage.

What varies is depth. USDA publishes the arithmetic in its annual budget justifications, and it repays careful reading:

Two misreadings are common and both should be resisted. It is not true that 92% of imported meat is never examined: every lot passes a real check. It is true that 92.3% of imported pounds were not selected for the deeper physical examination or laboratory test where pathogens, chemical residues, and species substitution are actually found. Those are different kinds of assurance, and a refusal statistic tells you little about the second one.

Note also that these are pounds, not shipments or lots. A refusal rate by weight is not a refusal rate by consignment, and neither is a detection rate.

A related caution applies to what refusals are for. A working paper by a USDA-affiliated economist analyzing more than 130,000 refusal records from FY2016 through FY2023 found shipping damage and invalid shipping marks dominating the reasons, with beef and veal refused most often and Australia and New Zealand the leading origins. That describes what the census-level checks catch. It does not establish that the sampled checks are failing, and it should not be read that way.

3. How USDA decides a foreign system is safe

Before any shipment arrives, a country must be found equivalent. FSIS recognizes 37 countries as maintaining inspection systems equivalent to the U.S. system, with more than 30 others seeking that status.

Ongoing verification runs on three tracks. On-site audits are scheduled by risk — high-priority countries within 24 months, medium within 30, low within 36 — with a floor of at least once every three years. FSIS completed audits in 13 countries in FY2024 and 16 in FY2025. Between audits, each country must submit by May 18 an updated Self-Reporting Tool, its certified establishment list, the prior year’s chemical residue program results with follow-up on violations, and the prior year’s microbiological sampling results. Port-of-entry outcomes feed back into the same assessment, and delisted foreign establishments are tracked.

This is more than an honor system and less than continuous supervision. In the years between audits, the assessment rests substantially on what the exporting government reports about itself, tested against what shows up at the dock.

The staffing behind it has been flat: 124 full-time equivalents in FY2024, 129 in FY2025, budgeted back to 124 through FY2027, against a 14.8% increase in import volume in a single year. That program also carries export certification, so this is context rather than proof that import capacity declined. But the two lines are moving in different directions.

4. What the auditors found — and stopped looking at

USDA’s Office of Inspector General has examined this system twice in the modern era. Both reports deserve reading before anyone declares the border either secure or broken.

In September 2017, OIG evaluated FSIS’s equivalency assessments of exporting countries. FSIS then sorted countries into three performance tiers — “adequately performing,” “average performing,” and “well-performing” — carrying audit obligations of annually, every two years, and every three years respectively. Of 31 eligible countries, 24 sat in the “adequately performing” tier, the lowest of the three, and none qualified as well-performing. OIG issued eight recommendations, and found that four of seven recommendations from a 2008 audit had not been adequately carried into the manual that replaced the agency’s guidance in October 2015.

In July 2020, OIG turned to the ports. Reviewing fiscal years 2017 through 2019 across nine reinspection establishments in three district offices and more than 300,000 imported lots, it found that seven of eight import inspection program personnel were verifying the label on a single carton per sample pallet rather than on every carton — in one instance verifying 0.7% of labels against a 10% requirement. Six of 53 examined shipments had lot counts that did not match the foreign inspection certificate, and five of those should not have been admitted. FSIS accepted all three recommendations.

Those are substantive findings with substantive corrective actions attached. What we cannot find is a successor. We located no OIG or GAO audit issued during 2021–2026 that revisits either equivalence determinations or port-of-entry controls. The Government Accountability Office did publish a significant meat and poultry oversight report in January 2025 — five recommendations, two of them designated priority, all five still open — but its scope is domestic pathogen standards and the outdated 2014 memorandum of understanding between FSIS and APHIS. It does not examine imports.

The consequence is that the two audits defining public knowledge of this system describe conditions from 2015 and from FY2017–2019. Import volume has grown by billions of pounds since.

5. Where the gap bites hardest: grass-fed

The origin gap is not evenly distributed. It falls heaviest on the claims shoppers pay a premium for.

USDA’s Agricultural Marketing Service withdrew its “Grass (Forage) Fed” marketing claim standard effective January 12, 2016, along with “Naturally Raised,” concluding that the standards did not usefully facilitate marketing and that the agency should defer to voluntary consensus standards developed outside government. Companies were told to adopt a private standard or write their own. There has been no federal grass-fed definition since.

Now combine three facts. There is no federal grass-fed standard. There is no retail country-of-origin requirement for beef. And industry analysis has long placed imported product at roughly 75–80% of labeled grass-fed beef sales in the United States — an estimate originating in a 2017 Bonterra Partners report, old enough now to deserve fresh measurement, but never seriously contradicted. The result is a package that can carry a company’s own grass-fed definition, describing cattle raised on another continent, with nothing on the label obliged to say so.

One thing did change. As of January 1, 2026, USDA’s revised “Product of USA” rule is enforceable: an unqualified “Product of USA” or “Made in the USA” claim on a meat, poultry, or egg product now requires that the animal was born, raised, slaughtered, and processed in the United States. That closes the false positive. It does not create a true positive. Imported grass-fed beef may no longer claim America — and still is not required to name the country it came from.

There is legislative movement. The American Beef Labeling Act of 2025 (S. 421) cleared the Senate Agriculture Committee 17–6 on August 6, 2026 during farm bill markup, and would reinstate mandatory country-of-origin labeling for beef. It is attached to a farm bill whose passage is uncertain, and it is not law.

6. What we could not verify

Three gaps, stated plainly so that no one builds on sand.

FY2021–FY2023 reinspection volumes. The detailed import table first appears in the FY2026 budget chapter, so the comparable series we can source runs only FY2024–FY2025. Earlier years exist in FSIS’s downloadable import datasets but are not published in this consolidated form.

Current audit dates by country. FSIS’s Foreign Audit Reports index does not respond to automated retrieval. We could not confirm the most recently published audit for Brazil or for other major suppliers, and we decline to assert a date we have not seen.

A leakage rate. Product does reach commerce without reinspection. “Import violation” was the leading FSIS recall reason in calendar 2023, at 24.6% of 65 recalls, and accounted for 5 of 42 recalls in 2025; in August 2026 a Florida importer recalled roughly 29,600 pounds of Argentine boneless beef distributed in Texas and Florida for exactly that reason. But no published denominator of shipments exists against which to compute a rate, so these are reported as a share of recalls and nothing more.

7. What this means for FAT’s 16 categories

Three of FAT’s sixteen disclosure categories are directly implicated, and all three fail in the same way — not for want of data, but for want of a requirement to pass it along.

Category 5, Farm / Country of Origin. The country is declared on the carton at the border under 9 CFR 327.14. For beef and pork it is not required on the retail package. The information exists in the import file and stops.

Category 6, Supply-Chain Intermediaries. The importer of record is a matter of public record on the vessel manifest — except where a lawful confidentiality filing removes the name, the mechanism documented in Purchased Silence. Between the foreign plant and the American shelf sit brokers, cold-storage operators, and further processors that no label names.

Category 7, Processor. The foreign establishment number appears on the imported carton. Once product enters a domestic establishment for further processing, the establishment number on the retail package is the American one, and the foreign plant that actually slaughtered the animal disappears from the consumer-facing record.

This is the distinction that should govern what we ask for. It is not a data gap. It is a disclosure gap: the government already requires, collects, and files the information at the border, and simply does not require it to travel the last hundred feet.

8. Recommendations

To FSIS. Publish the import reinspection table for FY2021–FY2023 in the same consolidated form used in the FY2026 and FY2027 budget chapters, so that a five-year trend can be evaluated by anyone. Publish, on a single indexed page, the date of the most recent completed equivalence verification audit for every eligible country.

To USDA OIG. Both prior audits of this system are now describing conditions from 2015 and FY2017–2019, across a period in which import volume rose sharply and international program staffing did not. A follow-up on the 2020 port-of-entry findings, and on the 2017 equivalence findings, is overdue.

To Congress. The disclosure gap closes with a requirement that origin already declared at the border be carried to the retail label. S. 421 is one vehicle; the principle does not depend on it.

To retailers and brands. Nothing prevents voluntary disclosure now. A brand that knows the country and the foreign establishment — and every importer does — can print them.

Sources

• USDA – FY2027 Explanatory Notes, Food Safety and Inspection Service
• USDA – FY2026 Explanatory Notes, Food Safety and Inspection Service
• FSIS Directive 9780.1 – Verifying the Ongoing Equivalence of Foreign Food Safety Inspection Systems
• FSIS Directive 9900.5 – Label Verification of Imported Meat, Poultry, and Egg Products
• 9 CFR 327.14 – Marking of products and labeling of immediate containers for importation
• USDA OIG 24601-0002-21 – Evaluation of FSIS’s Equivalency Assessments of Exporting Countries (Sept. 27, 2017)
• USDA OIG 24601-0003-21 – Controls Over Imported Meat and Poultry Products (July 1, 2020)
• GAO-25-107613 – Food Safety: USDA Should Take Additional Actions to Strengthen Oversight of Meat and Poultry (Jan. 2025)
• USDA ERS – Livestock, Dairy, and Poultry Outlook, LDP-M-381 (March 2026)
• Ahn, J.-W. – “Import Refusal Trends of FSIS-Regulated Products: Meat and Poultry” (working paper, July 2025)
• National Agricultural Law Center – Country of Origin Labeling (COOL): An Overview
• Federal Register – Withdrawal of U.S. Standards for Livestock and Meat Marketing Claims (Jan. 12, 2016)
• SDSU Extension – Market Share of Grass-Fed Beef (citing Bonterra Partners, 2017)
• American Beef Labeling Act of 2025, S. 421, 119th Congress
• USDA FSIS – Recall: Corte Argentino USA LLC, beef imported without the benefit of import reinspection (Aug. 11, 2026)
• Farm Animal Transparency – Understanding the 16 FAT Categories

Working paper. Figures verified against primary sources as of September 10, 2026. Where a figure could not be traced to a primary source it is identified as unverified in Section 6. Corrections welcome: dirkadams@farmanimaltransparency.com.

Companion research: Purchased Silence traces, by name and from public vessel manifests, which American beef companies receive imported frozen boneless beef — and which have purchased confidentiality to keep their names off the public record.