FDA Recall, Inspection and Warning Letter Statistics
Each finding below is one sentence you can quote as written, with its date. Every number comes from the query printed under it, so you can check it or rerun it. Use the permanent link next to a finding to cite it, and the embed code under a chart to republish the chart with its source credit.
Recall records, 2024 to 2025
FDA recall records fell 0.7% from 5,222 in 2024 to 5,187 in 2025, counting drug, device and food enforcement reports by recall start date, according to RegDataLab's analysis of FDA data as of October 31, 2026.
Permanent link to this finding · Related: Browse FDA recalls started in 2025 · FDA recall trends by year
Chart data
| Period | Recall records |
|---|---|
| 2016 | 7,294 |
| 2017 | 7,077 |
| 2018 | 6,671 |
| 2019 | 6,606 |
| 2020 | 4,649 |
| 2021 | 4,913 |
| 2022 | 5,093 |
| 2023 | 5,955 |
| 2024 | 5,222 |
| 2025 | 5,187 |
Embed this chart
Copy this HTML to publish the chart. Keep the source line and the link.
Query and source
Source: openFDA drug, device and food enforcement reports (table raw_enforcement; one record per recalled product).
SELECT EXTRACT(YEAR FROM recall_initiation_date)::int AS year,
COUNT(*)::int AS recalls,
COUNT(*) FILTER (WHERE classification = 'Class I')::int AS class_i
FROM raw_enforcement
WHERE recall_initiation_date >= DATE '2016-01-01'
AND recall_initiation_date < DATE '2026-01-01'
GROUP BY year
ORDER BY yearWhich products were recalled most in 2025
Medical devices accounted for 2,838 of the 5,187 FDA recall records started in 2025 (54.7%), followed by food (1,571) and drugs (778), according to RegDataLab's analysis of FDA data as of October 31, 2026.
Permanent link to this finding · Related: Browse all FDA recalls · FDA recalls by product type
Query and source
Source: openFDA drug, device and food enforcement reports (table raw_enforcement).
SELECT product_type, COUNT(*)::int AS recalls
FROM raw_enforcement
WHERE recall_initiation_date >= DATE '2025-01-01'
AND recall_initiation_date < DATE '2026-01-01'
AND product_type IS NOT NULL
GROUP BY product_type
ORDER BY recalls DESCInspections that ended in official action, FY2025
Of the 15,756 FDA inspections classified in fiscal year 2025, 605 (3.8%) were Official Action Indicated (OAI), meaning FDA recommended regulatory or administrative action, and 4,589 (29.1%) were Voluntary Action Indicated (VAI), according to RegDataLab's analysis of FDA data as of October 31, 2026.
Permanent link to this finding · Related: Browse OAI inspections · FDA inspection statistics dashboard · Check a supplier's inspections in Claude
Chart data
| Period | Of classified inspections |
|---|---|
| FY2016 | 3.6% |
| FY2017 | 3.4% |
| FY2018 | 2.4% |
| FY2019 | 3.1% |
| FY2020 | 3.5% |
| FY2021 | 5.1% |
| FY2022 | 4.0% |
| FY2023 | 3.2% |
| FY2024 | 3.7% |
| FY2025 | 3.8% |
Embed this chart
Copy this HTML to publish the chart. Keep the source line and the link.
Query and source
Source: FDA Data Dashboard inspection classifications (table fda_inspections; one row per inspection).
SELECT fiscal_year,
COUNT(*) FILTER (WHERE code = 'OAI')::int AS oai,
COUNT(*) FILTER (WHERE code = 'VAI')::int AS vai,
COUNT(*) FILTER (WHERE code = 'NAI')::int AS nai
FROM (SELECT fiscal_year::int AS fiscal_year,
CASE
WHEN classification_code IN ('OAI', 'VAI', 'NAI') THEN classification_code
ELSE CASE upper(btrim(regexp_replace(classification, '[[:space:]]+', ' ', 'g')))
WHEN 'OAI' THEN 'OAI'
WHEN 'OFFICIAL ACTION INDICATED (OAI)' THEN 'OAI'
WHEN 'VAI' THEN 'VAI'
WHEN 'VOLUNTARY ACTION INDICATED (VAI)' THEN 'VAI'
WHEN 'NAI' THEN 'NAI'
WHEN 'NO ACTION INDICATED (NAI)' THEN 'NAI'
END
END AS code
FROM fda_inspections
WHERE fiscal_year BETWEEN 2016 AND 2025) classified
GROUP BY fiscal_year
ORDER BY fiscal_yearMost cited Form 483 regulation, FY2025
FDA investigators recorded 14,003 Form 483 citations in fiscal year 2025, and the most frequent was 21 CFR 1.502(a) (FSVP assurances), cited 462 times (3.3%), according to RegDataLab's analysis of FDA data as of October 31, 2026.
Permanent link to this finding · Related: Browse FDA inspections and their citations · Form 483 citation statistics · What investigators cited most at drug sites
Query and source
Source: FDA Data Dashboard inspection citations (table fda_citations; one row per Form 483 citation).
WITH cited AS (
SELECT act_cfr_number, MAX(short_description) AS short_description, COUNT(*)::int AS citations
FROM fda_citations
WHERE fiscal_year = 2025
GROUP BY act_cfr_number
)
SELECT act_cfr_number, short_description, citations,
(SELECT SUM(citations) FROM cited)::int AS total_citations
FROM cited
WHERE act_cfr_number IS NOT NULL AND act_cfr_number <> ''
ORDER BY citations DESC, act_cfr_number
LIMIT 1Warning letters outside tobacco, FY2025
FDA's compliance actions data lists 456 warning letters outside tobacco in fiscal year 2025, while 11,727 more (96.3% of all warning letters) came from the Center for Tobacco Products, according to RegDataLab's analysis of FDA data as of October 31, 2026.
Permanent link to this finding · Related: Browse FDA warning letters · FDA warning letter statistics · Get a free supplier FDA snapshot
Chart data
| Period | Warning letters |
|---|---|
| FY2016 | 556 |
| FY2017 | 450 |
| FY2018 | 346 |
| FY2019 | 327 |
| FY2020 | 467 |
| FY2021 | 420 |
| FY2022 | 390 |
| FY2023 | 389 |
| FY2024 | 403 |
| FY2025 | 456 |
Embed this chart
Copy this HTML to publish the chart. Keep the source line and the link.
Query and source
Source: FDA Data Dashboard compliance actions (table fda_compliance_actions; one row per action).
SELECT fiscal_year,
COUNT(*) FILTER (WHERE center IS DISTINCT FROM 'CTP')::int AS non_tobacco,
COUNT(*) FILTER (WHERE center = 'CTP')::int AS tobacco
FROM fda_compliance_actions
WHERE action_type = 'Warning Letter'
AND fiscal_year BETWEEN 2016 AND 2025
GROUP BY fiscal_year
ORDER BY fiscal_yearSample and method
- Recalls are recall records from openFDA's drug, device and food enforcement reports, counted by recall start date (recall_initiation_date). FDA lists one record per recalled product, so one recall event can produce several records. 2025 is the latest calendar year that ended at least 90 days before the as-of date.
- Inspections, Form 483 citations and warning letters come from the FDA Data Dashboard and are counted by FDA fiscal year (October 1 to September 30). FY2025 is the latest fiscal year that ended at least 180 days before the as-of date, because FDA adds classified inspections and citations months after they happen.
- Inspection classification uses FDA's code (OAI, VAI or NAI). Inspections with no recognizable code are left out of the share.
- Warning letters are split by issuing center. Most warning letters in the dashboard come from the Center for Tobacco Products, so totals that include them describe tobacco enforcement, not drug, device or food manufacturing.
- Counts are recomputed hourly from the latest data refresh. Earlier years can change slightly when FDA corrects or adds records, so quote a finding with its as-of date.
- FDA data is a U.S. government work. Data freshness for each source is on the recall, inspection and compliance statistics pages. Machine-readable findings: /api/research.
How to cite
RegDataLab, "FDA Recall, Inspection and Warning Letter Statistics," October 31, 2026, https://www.regdatalab.com/research. Link a single finding with its anchor, for example https://www.regdatalab.com/research#inspection-oai-share.