The 15 days that decide whether you get a warning letter

An FDA Form 483 “notifies the company’s management of objectionable conditions” at the end of an inspection. It “does not constitute a final Agency determination,” and FDA encourages companies “to respond to the FDA Form 483 in writing with their corrective action plan” (FDA).

Timing matters. Since 2009, “FDA will generally allow firms 15 business days to provide a response to FDA 483 observations.” A response received in time is reviewed in detail before FDA decides on a warning letter. A late one usually is not: FDA “will not ordinarily delay the issuance of a warning letter” (Federal Register).

For what a Form 483 is and the observations FDA cites most, see FDA Form 483 Explained: What It Means, Top Observations and How to Respond.

What FDA cites most

AreaFY2025 top citations (system-generated 483s)
Devices (2,660 citations)§ 820.100(a) CAPA procedures: 279; § 820.198(a) complaints: 211; § 820.50 purchasing: 115; § 820.90(a) nonconforming product: 95; § 820.75(a) process validation: 93
Drugs (2,837 citations)§ 211.22(d) quality unit procedures: 243; § 211.192 investigations: 164; § 211.100(a) written procedures: 162

Source: GMP Insiders analysis of FDA data. FDA notes its spreadsheets cover only system-generated 483s and are “not a comprehensive listing of all inspectional observations” (FDA).

For devices, the rulebook changed on February 2, 2026. In the first ~100 QMSR inspections, the top topics were risk management, outsourcing and purchasing, complaint handling, UDI and corrective action (RAPS).

  • New draft guidance for drug sites. In March 2026, FDA issued draft guidance on “Responding to FDA Form 483 Observations at the Conclusion of a Drug CGMP Inspection” (Federal Register). It is a draft and covers drug CGMP inspections only. It recommends “one organized written response within 15 business days, rather than piecemeal submissions,” signed “by a member of executive management” (Venable), and cites past responses that failed “to address the root cause.”
  • More unannounced foreign inspections. In May 2025, FDA said it would “expand the use of unannounced inspections at foreign manufacturing facilities” (FDA).
  • AI-assisted targeting. FDA says its AI tool Elsa is used to “identify high-priority inspection targets” (FDA).

The FDA 483 response checklist

The structure below follows the elements law firms highlight from FDA’s 2026 draft guidance (Arnold & Porter; RAPS). It is written for drug sites but works as a sound structure for device firms too.

Day 0–2: Organize

  • Calendar the 15-business-day deadline from the inspection close
  • Assign an owner and a cross-functional team per observation
  • Assess immediate patient and product risk; decide on any holds or field actions

Day 2–10: Investigate

  • Investigate each observation to root cause, not just the specific example cited
  • Assess whether other products, lines or sites are affected (systemic scope)
  • Avoid “human error” as a final root cause without supporting evidence
  • Define corrections, corrective actions and effectiveness checks with dates

Day 10–15: Write and submit

  • One organized response, signed by executive management
  • Table of contents and a copy of the 483
  • Executive summary table: observation, root cause, actions, owner, target date, status
  • Objective evidence attached for completed actions (revised SOPs, training records, data)
  • Realistic timelines for open actions and a commitment to interim updates
  • Correspondence in English, with translations accompanied by originals

After submission

  • Send interim updates as actions close
  • Verify effectiveness and keep evidence ready for the next inspection

Common mistakes

MistakeWhy it hurtsBetter approach
Fixing only the cited exampleFDA reads it as a missed systemic issueShow the scope assessment across products and sites
”Retrained the operator” as the only actionTraining rarely addresses root causeShow the process or system change, then training
Promises without dates or evidenceCannot be verifiedDated commitments plus attached evidence
Late or piecemeal responsesFDA may not review before deciding on a warning letterOne complete response within 15 business days

Get your evidence straight before you write

A strong 483 response depends on finding every related procedure, record and prior investigation fast, then showing exactly how each was fixed. AI document assessment can map your SOPs and records against the cited requirement, flag related gaps the investigator did not list, and cite the passage behind every finding.

Run a sample regulatory document assessment on the SOPs behind one observation. AskDona shows the evidence; your team decides.

FAQ

Do I have to respond to a Form 483? Responding is voluntary, but FDA encourages a written response with a corrective action plan (FDA).

How long do I have? Generally 15 business days for FDA to consider the response before deciding on a warning letter (Federal Register).

Is a 483 a finding of violation? No. It “does not constitute a final Agency determination” (FDA).

Does the 2026 draft guidance apply to device makers? No. It covers drug CGMP inspections and is still a draft (Venable).

References

All pages opened and checked by October 9, 2026.

SourceTypeDateUsed for
FDA: Form 483 FAQRegulator2020What a 483 is
Federal Register: Review of Post-Inspection ResponsesRegulatorAug 11, 200915 business days
Federal Register: Draft 483 response guidanceRegulatorMar 9, 2026Draft guidance, root cause
Arnold & Porter: FDA guidance on 483 responsesLaw firmMar 12, 2026Recommended contents
Venable: Roadmap for respondingLaw firmJun 12, 2026One response, executive signature
RAPS: FDA drafts 483 response guidanceTrade pressMar 6, 2026Table of contents, language
GMP Insiders: FY2025 observationsIndustry analysisJan 5, 2026Top citations
FDA: Inspection observations datasetsRegulator dataDec 2025Dataset caveat
RAPS: Top QMSR observationsTrade pressMay 14, 2026QMSR topics
FDA: Unannounced foreign inspectionsRegulatorMay 6, 2025Foreign inspections
FDA: Elsa launchRegulatorJun 2, 2025AI targeting