21 CFR 211: CGMP for Finished Pharmaceuticals
Follow the path from approved procedures to a reviewable batch record, with clear product boundaries and practical evidence checks.

On this page · 12 sections
- 1What 21 CFR 211 covers—and how Subparts A–K fit together
- 2Subpart B: quality-unit authority and decision ownership
- 3Subpart C: connecting facility context to a batch
- 4Subparts D–E: equipment identity and material traceability
- 5Subpart F: approved procedures, execution, and deviations
- 6Subparts G–H: packaging evidence and distribution handoff
- 7Subpart I: laboratory results in their batch context
- 8Subparts J–K: the batch record and its release-review boundary
- 9Investigations: triggers, scope, conclusions, and followup
- 10Product and jurisdiction mapping without false equivalence
- 11Integrations: reconcile identities, revisions, and failed transfers
- 12Implementation and review checklist
How does 21 CFR 211 apply to your shop floor?
Pick your industry and scale — Ask V5 rewrites the definition in your context, gives a worked example, and shows what V5 does on day one.
01What 21 CFR 211 covers—and how Subparts A–K fit together
21 CFR 211 is Title 21, Part 211 of the Code of Federal Regulations, setting minimum current good manufacturing practice requirements for preparation of drug products for administration to humans or animals (21 CFR 211.1(a)).
The scope expressly excludes positron emission tomography drugs and medical gases as defined in § 213.3(b)(12), under 21 CFR 211.1(a). That exclusion should not be extended to other product categories by analogy.
This article uses “21 CFR Part 211” as the readable alias. It is not interchangeable with CGMP, the broader manufacturing-quality concept. Use the following map to find the relevant discussion, not as a substitute for individual provisions.
The subpart titles and section ranges below are taken from the official eCFR table of contents for Part 211; the third column is this article's own mapping, not regulatory text.
| Subpart (official title) | Sections | Connection to this article (editorial) |
|---|---|---|
| A — General Provisions | 211.1–211.3 | Scope and applicability |
| B — Organization and Personnel | 211.22–211.34 | Quality-unit authority and decision ownership |
| C — Buildings and Facilities | 211.42–211.58 | Facility context for a batch |
| D — Equipment | 211.63–211.72 | Equipment identity |
| E — Control of Components and Drug Product Containers and Closures | 211.80–211.94 | Material traceability |
| F — Production and Process Controls | 211.100–211.115 | Procedures, execution, and deviations |
| G — Packaging and Labeling Control | 211.122–211.137 | Packaging evidence |
| H — Holding and Distribution | 211.142–211.150 | Distribution handoff |
| I — Laboratory Controls | 211.160–211.176 | Results in batch context |
| J — Records and Reports | 211.180–211.198 | Batch records and review |
| K — Returned and Salvaged Drug Products | 211.204–211.208 | Return and salvage boundaries |
| Term | Role | Practical distinction |
|---|---|---|
| CGMP | Broad manufacturing-quality concept | Not the name of one regulation |
| 21 CFR Part 211 | Named regulation | Specific provisions with a defined scope |
| Master production/control record | Approved production reference | Instruction baseline; see the detailed record discussion below |
| Batch production/control record | Record for an individual batch | Batch-specific execution evidence; see the detailed record discussion below |
An electronic batch record is a way to organize that batch evidence; the record discussion below addresses its content and review boundary.
Four categories remain distinct throughout the article:
- Binding regulatory duties: requirements tied to cited provisions and their applicable scope.
- Guidance: separately identified recommendations, not presented as the regulation itself.
- Recommended operating practices: practical methods for organizing work and evidence, not additional legal duties.
- Software support: functions that assist execution or review, not proof that a site is compliant or a substitute for human approval and release.
02Subpart B: quality-unit authority and decision ownership
The quality control unit is not simply a testing laboratory. Under 21 CFR 211.22(a), it has responsibility and authority to approve or reject all components, drug product containers, closures, in-process materials, packaging material, labeling, and drug products. That paragraph also makes it responsible for approving or rejecting drug products manufactured, processed, packed, or held under contract by another company.
Under 21 CFR 211.22(c), the unit must approve or reject all procedures or specifications affecting drug product identity, strength, quality, and purity. This is decision authority over the standards and instructions governing the product—not just its test results. 21 CFR 211.22(d) requires the unit’s responsibilities and procedures to be written and followed.
As a governance practice, use an authority matrix to make these responsibilities workable across production, laboratory, quality, and contracted operations. For each decision, identify:
- The decision owner and the approval or rejection being made.
- The supporting records needed before a decision.
- Delegated preparation tasks, such as assembling results or checking completeness.
- The escalation route for missing records, disagreements, or unavailable approvers.
Preparing a review package does not itself confer approval authority. For contracted work, use a quality agreement to clarify interfaces rather than assume one universal owner–contractor allocation. Apply the same distinction when assigning change-control responsibilities: name who prepares the assessment separately from who decides. The matrix is a practical governance tool, not a format prescribed by these paragraphs.
03Subpart C: connecting facility context to a batch
Check the current text of Subpart C for facility-specific obligations; a records checklist does not replace those requirements. As a suggested traceability practice, record a stable room or area reference where it helps connect a batch to facility events, maintenance activity, or a relevant incident. This production context tagging supports assessment, rather than defining universal mandatory batch-record fields.
For example, assess a room incident against the operations and batches present during the relevant period, using event timing and actual exposure. Do not treat every batch ever assigned to that room as equally affected merely because it shares the location reference.
04Subparts D–E: equipment identity and material traceability
Equipment and material evidence answers two different questions: what was used, and how much? The following are batch-record duties under § 211.188, not a complete statement of Subparts D and E. Check those subparts’ current texts for their wider duties.
- 21 CFR 211.188(b)(2) requires identification of the individual major equipment and lines used.
- 21 CFR 211.188(b)(3) requires specific identification of each batch of component or in-process material used.
- 21 CFR 211.188(b)(4) requires recording the weights and measures of components used during processing.
Worked case: correct mass, incomplete identity
Assume an approved target of 20.000 kg of one component. The first addition is 12.000 kg; the second is 8.000 kg. Total added mass is 12.000 kg + 8.000 kg = 20.000 kg, exactly matching the target. However, the second addition has no component-lot identifier.
The arithmetic reconciles quantity, not provenance: it cannot establish which lot supplied the second addition. The traceability gap remains a separate review question. This example establishes neither an acceptable correction nor a batch disposition.
As a record-design practice, connect batch → material-use event → component lot, and execution step → equipment identifier. Link each material consumption posting to its actual use event so lot genealogy preserves identity alongside quantity. These relationships are a suggested implementation, not additional prescribed regulatory fields.
05Subpart F: approved procedures, execution, and deviations
A controlled procedure establishes what should happen; an execution record shows what actually happened. Keep those functions distinct when designing production records: approval of an instruction is not evidence that a particular batch followed it.
Under 21 CFR 211.100(a), written production and process-control procedures, including changes, must be drafted, reviewed, and approved by the appropriate organizational units and reviewed and approved by the quality control unit. In practice, connect change control to the procedure revision used for execution, rather than merely retaining the latest approved document.
21 CFR 211.100(b) requires written procedures to be followed during production and process-control functions and documentation at the time of performance. It also requires any deviation from written procedures to be recorded and justified. The separate investigation triggers and scope are addressed in the investigation section.
Make the instruction-to-execution connection visible
As a record-design practice, link each execution to its applicable procedure revision, performer, event time, relevant measurement with units, and exception reference where applicable. A batch execution history should let a reviewer distinguish three things:
- Intended instruction: assuming a procedure specifies mixing for 10 minutes, that duration is the target—not an observed result.
- Observed result: an execution entry reports 8 minutes and identifies the performer and event time.
- Documented departure: an exception reference connects the shorter duration to its recorded explanation and justification, without overwriting the observed result.
Practical evidence lifecycle
Procedure approval → input identification → execution → result reconciliation → exception handling → review → disposition.
This diagram is an illustrative workflow, not a regulatory sequence prescribed step by step. Reconcile recorded results against the applicable instruction; connect exceptions to the affected step; and carry unresolved questions into review. Exception handling may loop back to reconciliation before a disposition decision.
- Approve procedureEstablish the applicable revision and responsible roles.
- Identify inputsConnect component lots, equipment, and batch identifiers.
- Execute and recordCapture performed steps and supporting measurements.
- Reconcile resultsAssociate laboratory and packaging evidence with the batch.
- Resolve exceptionsDocument departures, investigate applicable discrepancies, and track followup.
- Review batchAssess the record and unresolved issues.
- Record dispositionDocument the quality decision before downstream handoff.
06Subparts G–H: packaging evidence and distribution handoff
Packaging evidence belongs in the batch record, not just in a separate line log. The batch-documentation requirements include:
- Area inspections: document inspection of the packaging and labeling area before and after use under 21 CFR 211.188(b)(6).
- Labeling controls: include complete labeling-control records, with specimens or copies of all labeling used, under 21 CFR 211.188(b)(8).
- Containers and closures: include a description of the drug product containers and closures under 21 CFR 211.188(b)(9).
As an operational practice, link this evidence to the same batch identity used in production and to the documented disposition communicated downstream. Keep the quality decision distinct from its transmission: a warehouse acknowledgment confirms receipt of a status message, not a new batch release decision. Record the disposition sent and the acknowledgment against that batch so logistics can distinguish an approved status from an unconfirmed handoff.
These batch-record elements do not define the full packaging or distribution control system. Check the current texts of Subparts G and H for their wider obligations.
07Subpart I: laboratory results in their batch context
Laboratory work needs both adequate facilities and records that connect results to the batch. Under 21 CFR 211.22(b), adequate laboratory facilities must be available to the quality control unit for its testing and approval or rejection activities. This is a facilities obligation, not simply a requirement to retain test reports.
Batch production and control records must include in-process and laboratory control results under 21 CFR 211.188(b)(5), and document any sampling performed under 21 CFR 211.188(b)(10). The practical task is to make those results understandable alongside the sampling and production activities they describe.
As an industry-practice model, connect the sample identifier, batch identifier, test or method revision, result value, unit, applicable specification reference, and result status. Establish those relationships during sample login and handoff, then preserve them through MES-to-LIMS integration. These are suggested evidence relationships, not an exhaustive list of mandatory laboratory fields.
A numerically acceptable result attached to the wrong batch remains an unresolved record problem: it does not establish a result for the intended batch. Likewise, a value carrying the wrong unit cannot support a dependable comparison with its specification. A passing flag should not substitute for complete review.
Check the current texts of Subpart I and 211.194 for detailed laboratory and laboratory-record duties. Release review follows in the next section; investigation duties have their own section.
08Subparts J–K: the batch record and its release-review boundary
21 CFR 211.188 requires batch production and control records for each batch of drug product produced. The batch record is therefore batch-specific evidence, not merely a reusable instruction set.
Completing the master-versus-batch distinction, 211.188(a) requires an accurate reproduction of the appropriate master production or control record, checked for accuracy, dated, and signed. The master supplies the intended instructions; the executed record connects those instructions to what happened in a particular batch.
211.188(b) requires documentation that each significant manufacturing, processing, packing, or holding step was accomplished. Under 211.188(b)(11), identify the people performing and directly supervising or checking each significant step. For a significant step performed by automated equipment under § 211.68, identify the person checking it. This is not a universal two-signature rule for every entry.
The batch-release boundary is explicit: 211.192 requires quality-control-unit review and approval of all production and control records, including packaging and labeling, to determine compliance with established, approved written procedures before release or distribution.
Illustrative evidence relationships
The following field names and relationships are suggested industry practice—not a mandated database schema or a description of V5 internals. Their purpose is to make the review path explicit.
| Relationship | Suggested fields | Review question |
|---|---|---|
| Batch → master revision | batch_id, master_revision_id | Which instructions governed this batch? |
| Batch → execution and material use | step_execution_id, material_lot_id | Which work and materials belong here? |
| Batch → tests and investigations | sample_id, result_id, investigation_id | Which findings inform the decision? |
| Batch → review decision | reviewer_id, decision, decision_time | Who recorded the decision, and when? |
As a practical review test, follow one significant step from its master instruction through execution evidence to the review decision. A collection of attachments is less useful if their batch relationships remain ambiguous.
For returned and salvaged products, check the current text of Subpart K before deciding disposition.
Batch Record
Connects product identity and the production reference.
- batch_id
- product_id
- master_record_revision
- batch_status
Step Execution
Connects performed work, equipment, timing, and people.
- step_execution_id
- batch_id
- equipment_id
- performed_at
- performed_by
- checked_by
Material Use
Connects each addition to its component lot.
- material_use_id
- batch_id
- component_lot_id
- quantity
- unit
Test result
Preserves sample identity, measurement meaning, and revision.
- result_id
- batch_id
- sample_id
- source_revision
- result_value
- result_unit
Investigation
Connects findings, affected batches, and followup.
- investigation_id
- linked_batch_ids
- conclusion
- followup_status
Quality Review
Records the reviewer and disposition decision.
- review_id
- batch_id
- reviewer_id
- decision
- decided_at
09Investigations: triggers, scope, conclusions, and followup
21 CFR 211.192 requires thorough investigation of any unexplained discrepancy, including percentage yield outside the maximum or minimum established in the master production and control records. Failure of a batch or any component to meet a specification also triggers investigation; this applies whether or not the batch has already been distributed. The investigation must extend to other batches of the same drug product and other drug products potentially associated with the specific failure or discrepancy. A written investigation record must include conclusions and followup. These are binding duties, not optional troubleshooting steps.
Worked case: yield below the master-record range
21 CFR 211.188(b)(7) requires statements of actual yield and percentage of theoretical yield at appropriate processing phases. Those recorded values support comparison with the applicable master-record bounds.
Assume theoretical yield is 100.0 kg, actual yield is 96.0 kg, and the approved master-record range at this phase is 97.0%–103.0%. These are illustrative limits, not regulatory tolerances.
- Percentage yield: 96.0 ÷ 100.0 × 100 = 96.0%.
- Distance below the lower bound: 97.0% − 96.0% = 1.0 percentage point.
- Difference from theoretical mass: 100.0 kg − 96.0 kg = 4.0 kg.
The calculation identifies a discrepancy, not its cause. For practical yield reconciliation, check weighing records, transfers, sampling, equipment events, and potentially associated batches before forming a conclusion. Do not equate the 4.0 kg difference with proven physical loss without supporting evidence.
Link records without confusing their purposes
As an operational practice, link a procedural-deviation record to the applicable investigation using shared batch and event identifiers. Recording the departure does not itself establish cause, affected scope, conclusions, or followup. Likewise, out-of-specification handling should connect the failed result to the investigation rather than substitute a status label for its reasoning.
The following are suggested investigation controls, not inspection-frequency claims.
| Failure mode | Why it matters | Suggested control |
|---|---|---|
| Unsupported cause assignment | A plausible explanation can conceal unresolved evidence. | Test the proposed cause against records and alternatives. |
| Affected-batch scope based only on convenience | Associated batches may be overlooked. | Document shared materials, equipment, timing, and inclusion/exclusion reasoning. |
| Conclusions without followup ownership | Actions may remain unfinished. | Assign owners, due dates, and completion evidence. |
11Integrations: reconcile identities, revisions, and failed transfers
A system-neutral integration map starts with record origins: planning or ERP supplies batch and material references; execution captures performed work; laboratory systems provide sample-linked results; equipment systems provide relevant equipment references; downstream logistics receives the documented disposition. These are possible architecture choices, not mandatory applications or interfaces. The review question is whether each transferred record retains its meaning and association with the batch.
For interfaces such as MES-to-LIMS exchanges, define a contract covering batch_id, source_record_id, source_revision, event_time, received_time, quantity, unit, and import_status. Specify identifier ownership, time-zone conventions, permitted units, and import-state meanings. Distinguish when work occurred from when a message arrived: late arrival must not be used to imply contemporaneous capture. Preserve source timing rather than substituting the receiving system’s timestamp.
| Failure | Review risk | Suggested recovery |
|---|---|---|
| Duplicate message | Duplicate apparent execution | Recognize the same source event |
| Unit mismatch | Misread quantity | Resolve the unit mapping |
| Revised laboratory result | Stale evidence | Retain revision context and reassess the affected review |
| Partial transfer | Incomplete batch view | Reconcile expected and received records |
| Wrong batch mapping | Misattributed evidence | Correct the association with traceable reasoning |
Use a bounded recovery sequence: detect the exception, isolate the affected import, reconcile against the source, make a controlled correction or replay, check for duplicates, and assess whether prior review decisions need reconsideration. Isolation should make the import’s unresolved status visible rather than leave ambiguous data available for routine review.
Keep correction reasoning and revision history available for audit trail review. Neither silent overwriting nor a successful retry demonstrates that the batch record is complete. Close the exception only after checking the expected record set and identifying affected review decisions. These are recommended interface controls, not a prescribed Part 211 system architecture.
12Implementation and review checklist
Use this practical checklist alongside the sections above, not as an additional regulatory standard. Keep binding duties, guidance, and site-selected controls distinct when assigning actions.
- Confirm product and market scope using the scope and product-mapping sections.
- Map cited duties to approved procedures using the Subpart F discussion; identify gaps and owners.
- Identify decision owners using Subpart B, including responsibility for contracted operations.
- Connect source records into a navigable batch execution history, following the Subparts J–K evidence model.
- Test material-lot and equipment traceability using the Subparts D–E worked case.
- Verify result identity, revision, and batch context using the Subpart I discussion.
- Assess unresolved exceptions against the investigations section; distinguish missing evidence from an explained discrepancy.
- Complete the documented review described under Subparts J–K; make its decision and supporting records retrievable.
- Confirm downstream acknowledgment of the approved status using the integrations section and the electronic release record.
Test a normal batch and deliberately incomplete scenarios: a missing component-lot identifier, revised result, or interrupted interface transfer. Can a reviewer find each issue and understand its effect without guessing? Treat this as a record-usability exercise, not regulatory validation or compliance certification.
Frequently asked questions
Q.Does calling a product a supplement, food, API, or medical device determine whether Part 211 applies?+
Treat the label as a starting point, not the applicability decision. 21 CFR 211.1(a) describes minimum CGMP for preparation of drug products for humans or animals, excluding PET drugs and specified medical gases. As a practical classification exercise, document intended use, product status, manufacturing stage, and any combination-product questions. Consult 21 CFR 210 and, where relevant, combination-product requirements rather than assuming a commercial category alone resolves the applicable manufacturing framework.
Q.Why is the food-related OTC provision not a general exemption for OTC drugs?+
21 CFR 211.1(c) states a conditional non-enforcement provision, not an exemption for every nonprescription medicine. It concerns OTC products whose products and all ingredients are ordinarily marketed and consumed as human foods, and which may also be drugs because of intended use. The provision directs application of food CGMP regulations, including 21 CFR 117. For an applicability assessment, document the food-marketing and consumption basis for both the finished product and every ingredient; OTC status alone does not establish those conditions.
Q.What should a quality review package contain when manufacturing is performed under contract?+
21 CFR 211.22(a) assigns the quality control unit responsibility for approving or rejecting drug products manufactured, processed, packed, or held under contract by another company. As an operational package, assemble the applicable batch record, test results, packaging and labeling records, deviations, investigation conclusions, and unresolved questions. Use the quality agreement to define transfer formats, timing, access to underlying records, and escalation contacts. Make missing or pending information visible rather than treating the contractor’s summary as the complete review package.
Q.Why does a passing certificate of analysis not settle the batch-release decision?+
A passing certificate addresses reported analytical results, not the entire manufacturing history. 21 CFR 211.192 requires quality control unit review and approval of production and control records, including packaging and labeling, before release or distribution. For batch release, the practical question is whether those records support an acceptable disposition, not simply whether reported tests passed. Reviewers should reconcile the certificate with the batch identity, applicable specifications, execution records, and any unresolved discrepancies before documenting their decision.
Q.How should an investigation team identify potentially affected batches that have already shipped?+
21 CFR 211.192 requires investigation regardless of whether the batch has been distributed and extension to potentially associated batches of the same drug product and other drug products. A practical search starts with the suspected failure mechanism: shared material lots, equipment, processing windows, methods, or storage conditions. Use lot genealogy to build the candidate population, then document inclusion and exclusion reasoning. Connect potentially affected shipped lots to distribution information for recall readiness; shipment status should not narrow the technical investigation.
Q.What should a reviewer do when an equipment identifier or significant-step performer is missing?+
21 CFR 211.188(b) calls for identification of individual major equipment and lines, and persons performing and supervising or checking significant steps; for qualifying automated steps, it identifies the checker. As review practice, flag the omission and seek attributable supporting records, such as equipment logs or batch execution history. Document any correction transparently, including its basis. Do not guess an identifier or present a later reconstruction as a contemporaneous entry. Escalate unresolved uncertainty for quality assessment before disposition.
Q.How should a procedural-deviation record connect to a separate investigation without duplicating the same narrative?+
21 CFR 211.100(b) requires procedural deviations to be recorded and justified; 21 CFR 211.188(b)(12) includes investigations under 211.192 in the batch record. A practical linked-record design keeps immediate facts, affected steps, and justification in the deviation record, with analysis and conclusions in the investigation. Use stable identifiers and reciprocal references, not copied narratives that can diverge. A process event log can organize chronology, while MES–QMS integration is a design consideration for maintaining those connections and visible investigation status.
Q.What additional scope questions arise when batch records and approvals are electronic?+
Check the current text of 21 CFR Part 11 for its applicability to the proposed records and signatures. As a practical system assessment, identify the authoritative record, hybrid paper/electronic boundaries, signer identity, signature meaning, access permissions, change history, retention, and retrieval arrangements. Determine how reviewers will see source data and linked investigations, not just a PDF summary. Treat software as support for controlled execution and review, not as compliance certification or a substitute for human approval and release.
Primary sources
Explore this topic
21 CFR 211 sits inside 2 overlapping topic clusters in our glossary. Every neighbour is one click away.
Master and executed records that prove a batch or device was made to spec.
Drug-product cGMP rules, ICH Q-series, and the regulators that enforce them.
Want to see how 21 CFR 211 could fit into your own records and workflows? Explore the related V5 pages or talk to our team about what applies to your operation.
