BPR (Batch Production Record): Requirements and Contents
Connect each batch to its manufacturing instructions, execution evidence, exceptions and documented human review.

On this page · 11 sections
- 1What a BPR records—and what it does not
- 2Choose the rule by product and jurisdiction
- 3Supplement BPR duties under 21 CFR 111.255
- 4Supplement BPR contents and supporting evidence
- 5Food records: organize around the food safety plan
- 6Connect execution to the right master-record version
- 7Run the lifecycle and recover integration exceptions
- 8Three worked cases: variance, food monitoring and missing results
- 9Review the evidence, then document the human decision
- 10Ten BPR failure modes to challenge before go-live
- 11Keep the record and its evidence retrievable
How does BPR 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 a BPR records—and what it does not
BPR (batch production record) is the batch-specific record of manufacturing activities and the information used to control and review that batch. Operationally, it connects the batch identity to what happened during manufacture: materials used, steps performed, observations, results and the people involved.
The central distinction is intended work versus actual work. Instructions describe what to do; an executed record documents what was done. A document can contain both, but blank instructions alone are not execution evidence.
| Term | Primary purpose | Relationship to execution | Boundary |
|---|---|---|---|
| BPR | Record a particular batch | Captures actual manufacturing activities and supporting evidence | More than a production schedule |
| Master manufacturing record | Define repeatable manufacturing instructions | Provides the intended method | Not proof that steps occurred |
| Pharma BMR | Document pharmaceutical batch manufacture | Common terminology for an executed batch record | Not a universally prescribed document title |
| Work order or batch traveler | Authorize, schedule or route work | May collect execution entries | Its title does not establish completeness |
| Certificate of analysis | Summarize test information | Communicates analytical results | Not the complete manufacturing history |
These are operational distinctions, not universal legal definitions. Local document names do not determine which regulatory framework applies.
A BPR need not be one physical folder. In paper, electronic or hybrid systems, a main document can link to controlled supporting records. Together, those records form the batch execution history; their identifiers and references should make clear which evidence belongs to the batch.
02Choose the rule by product and jurisdiction
Choose the applicable framework before designing the record. “BPR” is not a universal checklist: the named U.S. provisions address different products and record purposes. A shared template should preserve those differences rather than make every field appear legally mandatory.
| Product context | Provision | Record focus | Applicability boundary |
|---|---|---|---|
| U.S. dietary supplements | 21 CFR 111.255(a) | Prepare a batch production record every time a dietary-supplement batch is manufactured. | Supplement batch-record framework; detailed duties follow in the next section. |
| U.S. human food | 21 CFR 117.190(a) | Establish and maintain records documenting implementation of the food safety plan under that subpart. | Determine the subpart’s applicability separately from how a batch packet is organized. |
| U.S. drug products | 21 CFR 211.188 | Prepare batch production and control records for each batch, including complete production and control information. | Drug-product records; do not treat this provision as the supplement or food checklist. |
For related context, see the supplement BPR framework and 21 CFR 117 overview.
- Binding regulation: the duties attributed to the provisions above.
- Official guidance: identify agency recommendations separately from the regulatory duties.
- Industry practice: operational recommendations and illustrative schemas are design approaches, not additional regulatory requirements.
- Software support: tools can support the chosen record workflow; keep capability descriptions separate from legal duties and human approval responsibilities.
This comparison concerns only the named U.S. provisions. Manufacturers of APIs, devices or industrial chemicals, and manufacturers operating under other jurisdictions, should check their applicable requirements rather than transfer this checklist unchanged. As a practical design step, label each proposed field by its regulatory basis or operational purpose before adopting the template.
03Supplement BPR duties under 21 CFR 111.255
For dietary supplements, 21 CFR 111.255(a) requires a batch production record every time a batch is manufactured. Maintaining an unused template does not satisfy that per-batch duty: the template provides a structure, while the BPR records the particular batch. Repeated manufacture of the same formulation still triggers a record for each batch.
Under 21 CFR 111.255(b), the record must include complete information relating to both production and control of each batch. Treating it as only a production checklist misses the control side of that requirement. In practical record design, connect the account of what happened with the evidence used to assess it, rather than leaving reviewers to reconstruct those relationships.
21 CFR 111.255(c) has two linked duties: the BPR must accurately follow the appropriate master manufacturing record, and each production step must be performed. An orderly record alone is not a substitute for carrying out the work. The master manufacturing record provides the reference against which batch execution can be understood.
Three practical checks
Use these as an implementation approach, not as additional statutory wording or a prescription for electronic snapshot technology:
- Distinct batch identity: can a reviewer separate this execution and its supporting evidence from every other run of the same product?
- Identifiable master-record version: can the reviewer determine which instructions governed this batch, without relying on whichever version is current today?
- Evidence of actual execution: does the batch execution history distinguish completed work from planned, interrupted or unresolved work?
For an interrupted batch, preserve its actual state and identify outstanding work under site procedures. Do not populate completion entries merely to make the record appear finished. A useful handover tells the next operator or reviewer where execution stopped, what evidence already exists and what remains unresolved.
04Supplement BPR contents and supporting evidence
A supplement BPR should connect each recorded fact to evidence a reviewer can retrieve and interpret. The table focuses on core evidence relationships; check the current text of 21 CFR 111.260 for the complete contents list. The cited contents are legal requirements for supplement BPRs; the review questions are practical checks, not additional regulatory wording.
| Record element | Evidence to capture | Supporting reference | Review question |
|---|---|---|---|
| Finished-batch identity | The batch, lot, or control number of the finished dietary supplement batch. | 21 CFR 111.260(a)(1) | Does this identifier consistently connect the batch to its supporting records? |
| Equipment and processing lines | The identity of equipment and processing lines used in producing the batch. | 21 CFR 111.260(b) | Can the reviewer distinguish the actual equipment used from similar assets? |
| Equipment maintenance, cleaning and sanitizing | Dates and times of these activities for equipment and lines used, or a cross-reference to records such as individual equipment logs containing that information. | 21 CFR 111.260(c) | Does the reference locate the relevant entries, rather than merely the equipment-log collection? |
| Testing and examination | Results of any testing or examination performed during batch production, or a cross-reference to those results. | 21 CFR 111.260(h) | Can each referenced result be matched to the batch and the material examined? |
| Finished-supplement conformity | Documentation that the finished supplement meets specifications established under § 111.70(e) and (g). | 21 CFR 111.260(i) | Does the documentation support the stated conclusion for this finished batch? |
| Label documentation | An actual or representative label, or a cross-reference to its physical location, as specified in the master manufacturing record. | 21 CFR 111.260(k)(2) | Can the reviewer retrieve the actual or representative label identified by the record? |
| Component identity and amount | The unique identifier assigned to each component, packaging and label used, and the identity and weight or measure of each component used. | 21 CFR 111.260(d)–(e) | Does each component entry resolve to a received lot and a recorded weight or measure? |
| Yield | A statement of the actual yield and of the percentage of theoretical yield at appropriate phases of processing. | 21 CFR 111.260(f) | Is the yield stated at the phases the master record defines? |
| Monitoring results | The actual results obtained during any monitoring operation. | 21 CFR 111.260(g) | Are actual values recorded rather than a pass mark alone? |
| Step dates and initials | Documentation at the time of performance of the date each master-record step was performed and the initials of the persons performing it, including those weighing or measuring, verifying, adding and verifying addition of each component. | 21 CFR 111.260(j) | Can each step be tied to a date and to the people who performed and verified it? |
The specifications themselves are a separate topic; see finished product specifications. Here, the evidence relationship matters: a result records an observation, while conformity documentation connects the finished batch to the referenced specifications. Treat those as related records, not interchangeable field labels.
Implementation fields are not regulatory column names
Fields such as component_lot_id, quantity, unit, source_record_id and evidence_version are proposed implementation choices. They can make relationships explicit within a batch execution history, but the regulation does not prescribe those database column names. Define each field’s meaning so that an identifier is not mistaken for a result, or a quantity stored without its unit.
A useful cross-reference identifies a retrievable record, not merely a system name. For example, “laboratory system” is less useful than a report identifier and revision tied to the relevant sample and batch. For paper evidence, use a specific log, page or entry location. Test retrieval from the reviewer’s access level, not only from the record creator’s account. These reference-quality controls support traceability; the later review section addresses evaluation and the documented human decision.
05Food records: organize around the food safety plan
Following the scope distinction above, organize food records around the controls and activities they document. For operations within the relevant 21 CFR 117 scope, the required categories include:
- Preventive-control monitoring records under 21 CFR 117.190(a)(2).
- Corrective-action records and verification records under 21 CFR 117.190(a)(3)–(4). The listed verification categories carry the qualification “as applicable”; do not treat every category as relevant to every operation.
- Calibration of process-monitoring and verification instruments as a concrete verification category under 21 CFR 117.190(a)(4)(iv).
- Supply-chain program records under 21 CFR 117.190(a)(5).
- Applicable training records for the preventive controls qualified individual and qualified auditor under 21 CFR 117.190(a)(6).
Practical organization: link events, not duplicate packets
As an implementation practice, index evidence by control, location, time interval and affected lots. Use production context tagging to connect a record to the operations it actually covers, rather than assuming that its filing date establishes its batch relevance.
For example, assume one monitoring event covers a shared process during an interval in which two lots pass through it. Keep one event record and link both lots with their relevant processing times. A training record can similarly remain associated with the person, role and relevant operations without being copied into every batch packet.
These links make batch-level review useful without making the batch packet a substitute for the food safety plan’s record structure. Preserve broader-period and operation-level records in their own context.
06Connect execution to the right master-record version
With the master-record duty covered above, the implementation question is how to retain the instructions actually used. Bind each batch to master_record_id and master_revision. Preserve an instruction snapshot, or use another controlled reference that retrieves that exact revision of the electronic work instruction. These are configuration-control design patterns, not mandated field names, file formats or software architecture.
A revision authorized during execution is not necessarily the revision governing steps already completed. Record the authorized change decision, affected steps and reason; identify which instructions govern subsequent work. A recipe supersession log can track the transition, but should not silently replace historical instruction context. Keep completed actions linked to the instructions governing those actions.
The six-entity relationship diagram below is an illustrative evidence model, not an actual application schema. Arrows mean “links to”; a batch may have multiple related records.
| Relationship | Context retained |
|---|---|
| Batch → Step | Executed action and governing instruction revision |
| Batch → Material use | Lot, quantity and consuming step |
| Batch → Evidence | Result or document and source identifier |
| Exception → Step / Material use / Evidence | Affected records and reason |
| Review decision → Batch / Exception / Evidence | Reviewer, decision and evidence considered |
Preserve source identifiers and relationship keys in exports so batch execution history remains understandable outside the original interface. A display label alone may not distinguish otherwise similar results or actions.
Batch
Identifies production and instruction context.
- batch_id
- product_id
- master_revision
- status
Material Use
Connects consumed materials to production.
- material_use_id
- batch_id
- component_lot_id
- quantity
- unit
Step Execution
Records who performed each action.
- execution_id
- batch_id
- step_id
- occurred_at
- operator_id
Evidence Reference
Locates versioned supporting records.
- evidence_id
- batch_id
- record_type
- source_record_id
- version
Exception
Connects departures with their assessment.
- deviation_id
- batch_id
- affected_step_id
- disposition
Review decision
Preserves accountable human decisions.
- review_id
- batch_id
- reviewer_id
- decision
- signed_at
07Run the lifecycle and recover integration exceptions
Use this seven-step operating model to organize responsibilities and handoffs. It is not a sequence prescribed identically for supplements, food and pharmaceuticals; sequencing and ownership depend on the product, process and site procedures.
- Establish identity: associate the batch with its work order, product and material identifiers.
- Assign instructions: bind execution to the intended instruction version before work begins.
- Capture execution: record performed steps, quantities, operators and event times in the batch execution history.
- Link evidence: associate samples, results, equipment records and checks with the relevant batch or operation.
- Handle exceptions: identify missing, conflicting or duplicate evidence and assign recovery ownership.
- Review: assess execution, linked evidence and unresolved exceptions together, rather than treating a populated record as complete.
- Document disposition: have the authorized person record the batch decision under site procedures; successful data transfer is not approval.
Source responsibilities can differ. Planning or ERP may identify work orders and materials; execution systems may capture steps; laboratory systems may hold results; equipment logs may hold cleaning information. These are architectural examples, not promises of particular integrations. Define ownership at each boundary, including laboratory-result handoffs.
For recovery, reconcile expected evidence against received evidence using stable source IDs. Keep delayed results visibly pending, with an owner and follow-up action. Detect duplicate messages by source identity before adding evidence again; use controlled retries with logged outcomes. If a supporting record is unavailable, retain its reference and expose the gap rather than substituting a blank attachment.
Keep event time separate from interface receipt time: a late message does not prove late execution. Investigate inconsistent timestamps through clock-drift monitoring. Preserve the original record and correction reason, linking any replacement to what it supersedes.
- Identify batchAssign identity and confirm the applicable record type.
- Bind InstructionsConnect the batch to its authorized master version.
- Capture executionRecord materials, equipment, operators and observations during work.
- Connect evidenceLink logs, samples, results and relevant packaging records.
- Resolve exceptionsReconcile missing evidence and route exceptions for assessment.
- Review recordA designated person evaluates completeness and supporting evidence.
- Document dispositionAuthorized people record decisions and preserve the evidence.
08Three worked cases: variance, food monitoring and missing results
Case 1 — Supplement dispensing variance
Assume a planned quantity of 25.000 kg, an actual dispensed quantity of 24.750 kg and an illustrative site alert band of ±0.50% of planned quantity. Actual minus planned is −0.250 kg, or −250 g. Percentage variance is −0.250 ÷ 25.000 × 100 = −1.00%, outside the assumed alert band.
This is a dispensing comparison, not a yield calculation or a regulatory tolerance. The parameter tolerance band is site-defined for this example. The reviewer needs material identity, weighing evidence, the applicable instruction and the documented assessment. A percentage alone cannot establish whether the correct material was weighed or whether the recorded quantity represents the intended charge. Keep the original measurement linked to the assessment rather than replacing it with a corrected target value.
Case 2 — Food monitoring across two lots
Assume an illustrative site alert threshold of 8.0 °C, a measured value of 9.5 °C and two traceably affected lots. The exceedance is 9.5 − 8.0 = 1.5 °C. These temperatures demonstrate arithmetic; they are not recommended food-safety limits. Actual limits need process-specific justification.
Create one monitoring-event record linked to both lots, then link that event to the resulting assessment and corrective-action evidence. Each lot’s batch execution history can reference the same event without creating competing copies. Using the food-record categories covered above, the reviewer can follow the measurement, affected-lot relationship and response. The assessment should explain why both lots were included and record the conclusion for each; a shared event does not imply identical lot decisions.
Case 3 — An apparently complete laboratory import
Assume 120 expected result IDs and 120 delivered messages containing 118 unique expected IDs plus two duplicate deliveries. Two expected results are still missing. Receiving those two missing IDs produces 120 unique results across 122 delivered messages.
Reconcile the expected ID set against the received unique IDs, not just message totals. Preserve duplicate-delivery history while preventing duplicates from masquerading as additional results. An exception-based review can direct attention to the missing IDs and their recovery. A successful retry closes the transport gap; it does not establish that results are acceptable. Human review still determines what the recovered evidence means for the batch decision.
09Review the evidence, then document the human decision
For dietary supplements, 21 CFR 111.260(l)(1) requires documentation, at the time of performance, that quality-control personnel reviewed the batch production record. That review includes monitoring operations required under subpart E, under 111.260(l)(1)(i), and test and examination results, under 111.260(l)(1)(ii). The latter includes results for components, in-process materials, finished supplement batches, and packaged and labeled supplements.
The disposition itself is also a record requirement, not only a practice. 21 CFR 111.260(l)(3) requires documentation at the time of performance that quality control personnel approved and released, or rejected, the batch for distribution, including any reprocessed batch; 111.260(l)(4) requires the same for the packaged and labeled dietary supplement, including any repackaged or relabeled product. Paragraphs (m) and (n) add contemporaneous documentation of any required material review and disposition decision and of any reprocessing.
Separate completeness, acceptability and disposition
As an operational practice, ask three distinct questions: Is the record complete? Is the evidence acceptable? What is the authorized batch disposition? An assembled packet, completed batch release checklist, or electronic signature does not by itself answer all three. A result can be present yet unacceptable; a complete record can support rejection rather than release. People approve and release under their applicable responsibilities and procedures.
A practical reviewer sequence
- Confirm identity and instruction context: establish which batch, product and manufacturing instructions the evidence concerns.
- Reconcile expected evidence: distinguish a missing result from a result awaiting review, and identify any unexplained gaps.
- Examine exceptions and their impact: assess whether deviations, unexpected results or unresolved investigations affect the proposed disposition.
- Check supporting references: follow linked reports and logs to confirm that they concern this batch and support the recorded conclusions.
- Record the decision, decision-maker and unresolved restrictions: make the authorized status explicit in the disposition record or electronic release record. Do not leave readers to infer release from a review signature.
This sequence organizes batch review; it is not a product-independent legal release checklist.
10Ten BPR failure modes to challenge before go-live
Use the evidence relationships covered above as acceptance-test targets. These risk-based implementation challenges are industry practice, not an exhaustive regulatory audit checklist. Test whether a reviewer can reconstruct the batch without relying on undocumented explanations.
| Failure | Control to challenge |
|---|---|
| Wrong master version | Version-binding test: revise the master and confirm the existing batch retains its assigned version. |
| Mismatched batch identity | Identifier reconciliation: compare batch identifiers across execution, samples, results and packaging records. |
| Missing equipment reference | Retrieval test: follow the equipment reference to the relevant equipment log entry. |
| Ambiguous quantity unit | Unit and precision check: confirm recorded values retain units and intended decimal precision. |
| Duplicate result delivery | Unique-ID reconciliation: confirm repeated delivery creates no second logical result. |
| Missing step evidence | Expected-versus-recorded comparison: deliberately omit evidence and confirm the gap remains visible. |
| Overwritten correction | History inspection: use the audit trail review workflow to inspect original values, changes, authors and reasons. |
| Unlinked deviation | Affected-record linkage check: navigate from the deviation to every affected batch or step. |
| Unclear review ownership | Role and identity check: confirm each review task has an accountable role and each completed review identifies its reviewer. |
| Broken archive reference | Restoration test: restore the batch and resolve its supporting-record references, supporting recall readiness. |
Go-live checklist
- Name the owner of each evidence source.
- Establish the expected evidence set.
- Test an interrupted batch.
- Replay a duplicate interface message.
- Retrieve a shared supporting record.
- Restore an archived example.
A failed test calls for accountable remediation: assign an owner, assess impact, document the correction and retest. It is not an automatic legal conclusion about compliance.
11Keep the record and its evidence retrievable
For supplements under 21 CFR 111, 21 CFR 111.255(d) directs batch-production-record keeping to subpart P. For food, 21 CFR 117.190(b) makes the records required by that section subject to subpart F of 21 CFR 117. Check the current retention requirements for the relevant product, jurisdiction and record category; do not assign one retention period to every BPR.
As archive-integrity practice, preserve readable content, supporting-record references, version context and review history. A reference is useful only if the supporting record remains retrievable: keep identifiers and archive locations usable when applications, filenames or storage locations change.
Test retrieval after export or system replacement, not merely successful file creation. Select an archived batch, open its record and referenced evidence, and confirm that a reviewer can identify the applicable version and follow the review history. Record retrieval failures and resolve broken references before treating the archive as ready for routine use.
Frequently asked questions
Q.Can the same record template serve both a supplement line and a conventional-food line?+
Yes—as a template-design practice, use a common core with product-specific modules, rather than identical mandatory fields. For supplements, 21 CFR 111.255(a) requires a BPR whenever a batch is manufactured. For food, 21 CFR 117.190(a) addresses records documenting food safety plan implementation. Share identifiers, material entries and equipment references, but configure separate review checklists. Map the food module to applicable monitoring, corrective-action and verification records; do not treat a completed supplement template as proof that the food-record framework is covered.
Q.When is a linked laboratory report sufficient, and when is it useful to repeat key results in the batch record?+
For supplements, 21 CFR 111.260(h) permits testing or examination results, or a cross-reference to them. As a practical control, make the reference identify the report, sample, batch and report revision so reviewers retrieve the intended result. Repeat decision-critical values when this helps reviewers compare results with acceptance criteria, but label them as a summary and identify their source. A defined MES-to-LIMS handoff should address discrepancies between the summary and report rather than leaving reviewers to choose.
Q.How should an additional processing attempt be recorded without overwriting the first attempt?+
Use a separately identified attempt linked to the original operation. Preserve the first attempt’s settings, observations, quantities and outcome; record the reason, authorization, changed conditions and results of the additional attempt. For supplement reprocessing, 21 CFR 111.260(n) requires documentation at the time of performance, while 111.260(l)(2) requires documentation of quality control approval or rejection. As a practical design, the batch execution history should distinguish attempt-level outcomes from the final batch disposition, not replace an unsuccessful entry with a successful one.
Q.Can one equipment-cleaning log support several batch records?+
Yes, where each batch reference identifies the relevant equipment and cleaning event. For supplements, 21 CFR 111.260(c) permits a cross-reference to records retaining maintenance, cleaning and sanitizing dates and times. Practically, link to a specific entry rather than an entire logbook, and show the production sequence between cleaning events. Referencing one event from several batches does not itself establish that the cleaning remained suitable throughout the campaign; assess that separately through the site's sanitation and cleaning schedule and procedures.
Q.What should happen if a laboratory result changes after the batch record has been reviewed?+
Treat the change as a new review event, not a silent replacement. Preserve the earlier report and review, identify the corrected result, and route the change to the responsible quality reviewer for an impact assessment. Supplement 21 CFR 111.260(l)(1)(ii) requires documentation of quality control review of test and examination results. As a practical response, reassess affected conclusions and disposition; if product has shipped, evaluate downstream action. An audit trail review workflow can structure that assessment without making the decision.
Q.How should a contract manufacturer and brand owner arrange record access and review handoffs?+
As a contracting practice, define the review package and access arrangements before production starts. Specify who maintains originals, which supporting records are accessible, how corrections are communicated, and who acknowledges receipt. Separate the manufacturer's production review from any brand-owner acceptance step, with named decision-makers and escalation contacts. Include access to laboratory reports, deviations and subcontracted activities, not just a signed cover sheet. A shared batch release checklist can track these handoffs; contractual labels should not be treated as a determination of legal responsibility.
Q.How should a hybrid paper-and-electronic batch record identify which record controls when copies disagree?+
As a hybrid-record practice, define authority by record element: for example, signed paper observations versus electronically captured instrument results. Maintain an index identifying each authoritative record, its location and how copies are marked. When values disagree, preserve both, investigate the cause and document the resolution before relying on the disputed value. Do not assume that the latest timestamp or electronic format wins. A paperless shop-floor transition should explicitly manage these boundaries while both recording methods remain in use.
Q.How should an unfinished or abandoned batch be documented without implying that unperformed steps were completed?+
As a record-design practice, distinguish completed, started-but-incomplete, not-started and not-applicable steps. Record the stopping point, reason, material quantities and locations, equipment status, and the person making each entry. Preserve actual observations without filling expected values into unperformed steps. Link the batch to its hold, disposal or other disposition decision, and keep administrative closure separate from production completion. Contemporaneous recording helps preserve what actually happened; an abandoned batch should remain reconstructable even when it never reaches a finished-product stage.
Primary sources
Explore this topic
BPR sits inside this topic cluster in our glossary. Every neighbour is one click away.
Master and executed records that prove a batch or device was made to spec.
Want to see how BPR could fit into your own records and workflows? Explore the related V5 pages or talk to our team about what applies to your operation.
