An event identifier and a recall number are not the same thing. FDA guidance describes the recall number as an alphanumeric product-level tracking designation assigned after classification, while event-level tools group a firm's recall of one or more products into a single event [2][4]. Researchers who need to answer questions about scope, timing or status therefore must first decide whether the question concerns a classified product record or the broader recall event. The working method below separates those two units of analysis, walks through the available primary sources, identifies the technical decisions that affect results, and concludes with a suggested evaluation workflow grounded only in documented FDA materials.
Separate product recall numbers from event identifiers
FDA's Investigations Operations Manual distinguishes the recall number from the RES event ID. A recall number is assigned by the responsible center for each recalled product it classifies, and it consists of a center letter code, a 3- or 4-digit sequential fiscal-year product count, and a 4-digit fiscal year; the manual offers F-100-2024 as the 100th product classification of fiscal year 2024 [1]. The event ID is a different field that FDA display guidance lists beside recalling firm, status, recall number, reason and classification date [2].
That product-to-event relationship matters whenever a single firm action covers multiple products, lots or commodities. The FDA dashboard note states that an event is a firm's recall of one or more products, and that filtering at the event level shows how many recalls occurred while filtering at the product level shows how many individual products were recalled [4]. A researcher seeking a count of a manufacturer's recall activity will therefore get one number at event level and a larger number at product level, without any change to the underlying source records.
Where the identifiers come from and how to retrieve them
The public-facing recall data originates in the FDA Recall Enterprise System, which FDA describes as the electronic data system used by recall personnel to submit, update, classify and terminate recalls [8]. The openFDA drug and device enforcement APIs return data from RES and cover publicly releasable records from 2004 onward, updated weekly, with original records annotated and converted into JSON [3][7]. The Enforcement Report user interface offers weekly reports, advanced search, archived reports and an API, with an alternate Event View that groups information by recall event rather than by product [2].
A second retrieval path exists in the Recalls Data Sets page, which distributes XML files derived from firm-issued press release recalls from 2009 through 2026 [6]. These files are a different corpus from the weekly Enforcement Report: they reflect materials recalling firms issued, not the complete RES record set. Researchers combining both should record the source of every identifier because the same product action can appear in a firm press release, in a later Enforcement Report entry, and in RES-derived JSON without carrying identical fields or update timing.
Suggested evaluation workflow for an unfamiliar recall event
One workable sequence, offered as a suggestion rather than a tested protocol, is to start with a product-level recall number when one is available, note the responsible-center letter and fiscal year, then open the Event View for the same record to capture the event ID and the full set of commodities grouped under that event [1][2][7]. Next, compare the product-level count against the event-level count on the FDA dashboard, which displays only recalls classified on or after June 8, 2012, and note that older classified recalls are out of that dashboard's range [4].
Then record the classification state. If the entry is not yet classified, document that FDA has determined the action meets the recall definition but has not completed hazard assessment, and do not quote a class [5]. Finally, check for a firm-issued press release XML file and for any update history, while accepting that status after classification is not maintained in the public data [6][5][3]. This sequence keeps each identifier attached to the source that actually documents it.
Technical filters that change what a search returns
The search interface exposes fields including recalling firm, status, recall number, reason for recall, classification date and event ID [2]. One consequential filter is recall class. For recalls determined to meet the definition of a recall but not yet assessed, users can select "Not Yet Classified" from the recall class menu; that selection disables classified date fields because those records lack a classification date [2]. A date-bounded search that ignores not-yet-classified records can therefore silently exclude events that FDA has already determined qualify as recalls.
Classification status also changes what an identifier is evidence of. FDA states that all monitored recalls are included in the Enforcement Report once classified, and may be listed before classification when FDA determines the firm's action meets the definition of a recall; the entry is updated with the classification after hazard assessment [5]. A researcher viewing an event ID on a not-yet-classified record can confirm that FDA recognized a recall action, but cannot cite that entry as a hazard classification of Class I, II or III.
What the data cannot establish about status or lifecycle
OpenFDA includes an explicit disclaimer that the endpoint data should not be used to track the lifecycle of a recall, and that FDA does not update recall status after classification, so a status shown as open, completed or terminated remains unchanged after publication [3][7]. This means a completed status visible today is not a later-stage confirmation that a recall was closed; it is the status at the classification publication point. Researchers who need a current lifecycle determination should treat the Enforcement Report as a historical record, not an operational tracker.
Corrections can still occur. OpenFDA notes that FDA will make corrections or changes to previously disclosed recall information when necessary, for example if a firm discovers that the initial recall should be expanded to include more batches or lots [7]. The Enforcement Report also documents an update history feature for recalls posted or updated beginning July 24, 2018, covering fields such as classification, reason and product description when changes occur after initial publication [5]. Any research workflow should therefore re-query the record's history before treating an identifier's associated details as final.
Sources & references
- FDA Investigations Operations Manual 2026 — Chapter 7 Recall Activities · www.fda.gov
- Enforcement Report Information and Definitions | FDA · www.fda.gov
- Drug Enforcement Overview - openFDA · open.fda.gov
- FDA Dashboards - Recalls · datadashboard.fda.gov
- Enforcement Reports | FDA · www.fda.gov
- Recalls Data Sets | FDA · www.fda.gov
- Device Enforcement Overview · open.fda.gov
- FDA Recall Enterprise System (RES) · open.fda.gov