99 FDA Food Recalls: The USTA Recall Index

The USTA FDA Recall Index sealed 99 distinct food enforcement actions in its June 25, 2026 snapshot, drawn from the FDA's openFDA food enforcement feed. This feed covers FDA food recalls only — not USDA/FSIS meat-and-poultry recalls, not drug or device recalls, and not every U.S. food recall. Each of the 99 entries is a distinct recall_number, counted once using its most recent observation.

The most striking feature of this edition is the severity split. Class I and Class II are exactly tied at 47 recalls each — 47.5% apiece — with just 4 Class III recalls (4.0%) at the low end. That perfect equilibrium between the FDA's two highest urgency tiers defines what compliance professionals are working with in this sealed snapshot.

Class I and Class II each hold 47.5% of 99 sealed FDA food recalls.

Key Findings

• The sealed snapshot holds 99 distinct FDA food recalls as of June 25, 2026, according to US Tech Automations' sealed FDA-enforcement snapshot — drawn from the openFDA food enforcement feed only, not USDA/FSIS or drug and device feeds.

• 47 recalls are Class I (47.5%): the FDA's highest severity tier, defined as recalls with a "reasonable probability of serious health consequences or death."

• 47 recalls are Class II (47.5%): exactly tied with Class I, defined as recalls with "temporary or medically reversible health consequences."

• 94 of 99 recalls were initiated within the past 90 days, showing the feed is composed almost entirely of recent enforcement actions.

• Florida leads all states with 23 recalls (23.2%), followed by California at 22 recalls (22.2%) and Wisconsin at 17 recalls (17.2%) — a geographic concentration that reflects where the country's food manufacturing and processing operations are registered.

• 92 recalls (92.9%) carry Ongoing status: active enforcement is the overwhelming norm in this snapshot, not the exception.

The Recall Index at a Glance

An FDA food recall is a voluntary or FDA-requested action to remove or correct a product that violates FDA regulations. The recall_number is the primary identifier for each distinct enforcement action in the feed; our clock deduplicates to that identifier so a recall re-published daily as its status updates is counted exactly once.

The 90-day recency figure is instructive: 94 of the 99 distinct recall_numbers were initiated within that window. The handful of older entries — reaching back to December 15, 2025 — are active enforcement actions that have remained open past a typical resolution cycle, not historical artifacts.

Severity: A Perfectly Divided Classification Landscape

The FDA assigns every recall a classification based on the health risk associated with the violative product. There are three classes. Their meanings below are the FDA's own structured labels, quoted verbatim — not our assessment of any specific product:

Class I and Class II are exactly tied at 47 recalls each — a 47.5%/47.5% split that places the vast majority of active FDA food enforcement in the agency's two highest urgency tiers, with just 4 Class III recalls at the lower end.

The 47/47 Class I and Class II tie is the single most unusual feature of this snapshot. In most enforcement windows, one severity tier outnumbers the other by a meaningful margin. Here, the count is identical — meaning that for every recall the FDA classifies as carrying potential for serious or fatal harm, there is an equal count classified as likely reversible. That is not a signal that the recall landscape is safe; it is a signal that it is broad, spanning the full range of the FDA's two active severity tiers simultaneously.

For a food-safety compliance team, the operational question is not just "how many recalls are active" — it is "how many are Class I right now, and do any of them touch our supply chain." Here, the answer to the first part is 47. Unlike forward-looking data like the permit prediction ledger, which seals forecasts for later scoring against outcomes, the recall index seals current enforcement state: what the FDA is acting on, at what severity, at a specific sealed moment.

The sealed feed holds 47 Class I and 47 Class II FDA food recalls.

The 4 Class III recalls represent a thin but real category: the FDA found a violation, initiated an enforcement action, and classified the risk as "unlikely to cause adverse health consequences." These are real regulatory events — the classification reflects the FDA's own risk assessment, not an inference from the product description, which is not part of our sealed display set.

Status and Geographic Concentration

Enforcement Status Mix

Recall status tells you where in the enforcement lifecycle a given action sits. The feed carries three distinct status labels in this snapshot:

The Ongoing figure — 92 of 99 recalls — is the defining characteristic of this snapshot's lifecycle composition. An Ongoing recall is an unresolved enforcement action: the recalled product has not yet been fully removed, corrected, or otherwise closed out to the FDA's satisfaction. Only 5 have reached Completed status, and 2 were Terminated without reaching completion.

With 92 of 99 recalls carrying Ongoing status, this sealed FDA food-enforcement snapshot reads as a live enforcement landscape — not a historical record. Almost every entry is an unresolved action still in motion on the snapshot date.

Terminated differs from Completed: a Terminated recall means the action was ended before the standard recall process concluded. That may reflect a finding that the product was never actually distributed, or a determination that the recall was no longer necessary for another reason. The FDA's status labels are structural fields, and we do not infer the reason from free-text product descriptions.

Geographic Distribution

The state column in the FDA enforcement feed records the location attributed to each recall — typically the state where the recalling firm is registered or operates. The full distribution in this snapshot:

Florida (23 recalls, 23.2%) and California (22 recalls, 22.2%) lead — consistent with the scale of food manufacturing and distribution infrastructure in both states. Wisconsin's position at 17 recalls (17.2%) is more distinctive: it ranks third, ahead of New York's 7, reflecting the density of dairy, specialty food, and food-processing facilities concentrated in that state.

The geographic tail is thin. Six states or territories each contribute a single recall. One entry carries no state attribution at all — labeled "N/A" in the feed and counted here as "unspecified" since it does not represent a geographic location. State attribution is not a proxy for where consumers were affected; it is where the firm is registered. A compliance buyer using this data to monitor supplier geography should treat it as a location-of-origin field, not a distribution footprint.

This kind of regional signal — where concentration sits across a national dataset — is the same kind of structural read we produce in the MISO interconnection queue report, which maps infrastructure requests by region across the grid queue.

Methodology

The US Tech Automations FDA-enforcement clock pulls the FDA's openFDA food enforcement endpoint on a daily schedule and seals each snapshot with a content hash, making each day's observation independently verifiable. The source is the openFDA food enforcement feed — not USDA/FSIS (which covers meat and poultry separately), not the drug enforcement feed, and not the device recall database. This index covers the food feed only.

How the sealed recall index is built:

1. Pull. The FDA-enforcement clock queries the openFDA food enforcement endpoint daily, capturing all records available at that moment. The snapshot for this edition was sealed on June 25, 2026.

2. Deduplicate. Records are deduplicated to distinct recall_numbers. The unit is a distinct FDA recall_number; the most recent observation of each recall is used, so a recall re-published daily as its status updates is counted once.

3. Label. Classification (Class I / Class II / Class III) and status (Ongoing / Completed / Terminated) are the FDA's own structured labels, preserved verbatim and ungrouped.

No recall reason or severity is inferred from the free-text product description.
4. Seal. The deduplicated dataset is content-hashed and stored as an append-only snapshot. The content hash for this edition is recorded in this post's frontmatter.
5. Aggregate. Counts, share percentages, and geographic breakdowns are computed from the sealed distinct-recall_number dataset. Nothing is estimated or modeled.

Honesty statement: Every count is a verbatim count of distinct FDA recall_numbers the clock actually captured; nothing is estimated or modeled. Classification (Class I/II/III) and status are the FDA's own structured labels, ungrouped. No recall reason or severity is inferred from the free-text product description. Each recall is counted once using its most recent observation. Shares are over the distinct recalls captured in this feed, which is not the universe of all U.S. food recalls.

Scope statement: Food recalls and enforcement actions in the U.S. FDA's openFDA food enforcement feed, as captured and sealed daily by the US Tech Automations FDA-enforcement clock, deduplicated to 99 distinct recall events as of June 25, 2026. This covers the FDA food enforcement feed only — not USDA/FSIS meat-and-poultry recalls, and not drug or device recalls.

The sealed-snapshot discipline here is the same methodology we apply across our research indexes — from the USTA AI Price Index, which seals daily model-pricing data from the AI marketplace, to permit activity reports that seal public permit feeds for market intelligence use. Each index is built to be independently verifiable against a content-hashed snapshot, not re-computed on demand.

92 of 99 FDA food recalls in the sealed feed carry Ongoing enforcement status.

Frequently Asked Questions

Q: What is a Class I recall, and how does it differ from Class II?
A: A Class I recall is defined by the FDA as one where there is a "reasonable probability of serious health consequences or death" associated with the violative product. Class II means "temporary or medically reversible health consequences." Class III — just 4 recalls in this snapshot — means the FDA judges the product "unlikely to cause adverse health consequences."

These definitions come verbatim from the FDA's own classification system; we quote them as-is and make no additional inference about the severity of any specific product.

Q: Does this index cover meat, poultry, drug, or device recalls?
A: No. The USTA Recall Index is built exclusively from the FDA's openFDA food enforcement feed. Meat and poultry recalls are administered by USDA/FSIS, which maintains a separate recall system. Drug and device recalls appear in separate FDA enforcement feeds. This index covers food only and should not be read as a complete picture of all U.S. food safety enforcement actions.

Q: Why does the count land at 99 — could there be missing records?
A: The 99 figure is a verbatim count of distinct recall_numbers captured and sealed by the clock on June 25, 2026. It is not rounded or modeled. The openFDA food enforcement feed reflects what the FDA has published; records not yet published, or published after the snapshot was sealed, would not appear in this edition. The sealed count is what was observable on that specific date, and the content hash makes that count verifiable.

Q: What does "Ongoing" status mean for a recall in this feed?
A: Ongoing means the FDA enforcement action has not yet been resolved. The recalled product has not been fully removed from commerce, corrected, or otherwise closed out to the FDA's satisfaction. With 92 of 99 recalls carrying Ongoing status in this snapshot, the overwhelming majority of entries represent active, unresolved enforcement — not historical events. A Completed recall signals that the FDA has accepted the removal or correction; a Terminated recall means the action was ended before the standard process concluded.

Q: Why is Wisconsin ranked so high in the geographic breakdown?
A: Wisconsin's 17 recalls (17.2%) reflect the concentration of food manufacturing and processing operations registered in that state — particularly dairy, packaged foods, and specialty food producers. State attribution in the FDA enforcement feed corresponds to where a firm is registered or operates, not necessarily where consumers purchased or were affected by the recalled product. No additional inference about recall cause or product type is drawn from the state field or free-text product descriptions.

Q: How does a sealed snapshot differ from querying the FDA website directly?
A: The openFDA feed is a live dataset: querying it today returns a different result than querying it yesterday, as recalls are added, updated, or resolved. This index seals a specific daily observation with a content hash, making the June 25, 2026 count of 99 distinct recalls permanently verifiable and auditable. The same sealed-ledger approach governs the Los Angeles building permit report and our other research indexes — the point is that the snapshot is fixed at a known moment, not recalculated on each visit.

Put Recall Data to Work

FDA food recall monitoring is not a one-time read for anyone in the food supply chain. The 99 distinct recalls sealed here are a point-in-time baseline; the compliance and operational value lies in detecting movement from that baseline — new Class I additions, status changes from Ongoing to Completed, new entries from a supplier state you carry. Three recurring workflows illustrate who benefits from this data on a cadenced basis.

Food-safety and quality compliance manager. A compliance manager at a food manufacturer, processor, or co-packer monitors the FDA feed to catch new recalls before they surface in customer complaints, audits, or regulatory correspondence. With 47 Class I recalls in this snapshot, the relevant workflow is: compare each day's sealed count against the prior day, isolate any new Class I recall_numbers, and triage them against the firm's supplier and ingredient register.

That triage can run daily or weekly, but the escalation trigger is unambiguous — any new Class I entry warrants immediate review because the FDA definition ("reasonable probability of serious health consequences or death") makes delay costly. A compliance team monitoring 92 Ongoing recalls needs a structured alert, not a manual daily spot-check.

Grocery and retail category buyer. A category buyer at a regional chain or food retailer manages supplier relationships across dozens of SKUs. The geographic concentration in this snapshot — Florida at 23.2%, California at 22.2%, Wisconsin at 17.2% — maps directly onto where a supplier base is typically registered. Monitoring the recall feed by state lets a buyer flag risk at a supplier level before product reaches shelf. A state-filtered alert built on sealed data is more reliable than an ad-hoc search, and a Wisconsin-heavy supplier roster makes the 17 recalls from that state immediately actionable as a screening input.

Food-distributor operations manager. A distributor moving product between manufacturers and retail or foodservice accounts sits between the recall event and the end consumer. For this audience, status movement matters as much as classification: a recall shifting from Ongoing to Completed signals that the FDA has accepted removal is complete, while the 2 Terminated entries in this snapshot carry a different operational meaning. An operations manager needs a daily status-change alert keyed to the recall_numbers of products currently in their system — not a static snapshot checked quarterly.

The platform automates that monitoring — scheduled pulls from the FDA food-enforcement feed, alerting on new recall_numbers or status changes against your supplier and product register, and routing those signals into the compliance, buying, or operations workflow that needs them. See how the platform handles recurring data monitoring for food and supply chain teams.

Source: US Tech Automations Research — computed from the sealed daily FDA food-enforcement snapshot, June 25, 2026.