recalldb
Menu

Recall data downloads

For research, journalism and analysis. Free, with attribution.

On the first of each month we save a snapshot of every recall notice we hold: one row per notice, as gzipped CSV and JSON. Each row has the recall id, agency, category, who issued it, the recall date, a short title, units affected when the agency lists them, the official id and official link, a link to the plain-language page, and the data licence. They are summaries with links, not the full notices: the official notice is always the source.

How to get them

  1. Get a free API key (one per verified email).
  2. List the snapshots: GET /v1/downloads with Authorization: Bearer rdb_....
  3. Download one: GET /v1/downloads/recalls-2026-11.csv.gz (or .json.gz). Each download counts as one request against your 1,000 free requests a month and the usual rate limits.
curl
curl -H "Authorization: Bearer $RECALLDB_KEY" https://recalldb.app/v1/downloads
curl -H "Authorization: Bearer $RECALLDB_KEY" -o recalls.csv.gz https://recalldb.app/v1/downloads/recalls-2026-11.csv.gz

Licence and attribution

  • United States: NHTSA, CPSC, FDA and USDA FSIS records are U.S. government works in the public domain.
  • Canada, UK, Australia: Each row carries its licence (Open Government Licence - Canada, Open Government Licence v3.0, Creative Commons Attribution 4.0). Keep the attribution that licence needs.
  • RecallDB: Say "Data from RecallDB" with a link to https://recalldb.app wherever you publish results built on it.

Snapshots are only for lookup, research and reporting. Do not republish the files as a competing recall service. A recall notice does not say whether a specific item is affected, and a count of notices is not a measure of how good or bad a company is. See the terms.