Master Batch Record (MBR)

Master Batch Record (MBR) – Authoring, Control, and Execution Binding

This topic is part of the SG Systems Global regulatory & operations glossary.

Updated October 2025 • Records & Execution • MES, LIMS, ELN, Lot Traceability

The Master Batch Record (MBR) is the canonical, quality-approved recipe that dictates exactly how a batch is to be manufactured—no more and no less. It is not a static PDF that lives in a shared folder; it is an operational contract that binds together Planning, Production, Quality, Engineering, and Supply Chain by defining materials, quantities, tolerances, step sequencing, equipment and status requirements, in-process controls (IPC), sampling plans, labeling control, and release criteria with enough specificity that ambiguity is engineered out of day-to-day decisions. When authored properly and executed through an electronic batch record (eBMR) inside a modern MES, the MBR becomes the backbone of end-to-end genealogy and compliant release. When handled casually, it becomes a primary driver of deviations, rework, hold backlogs, and painful audit findings. In simple terms: the MBR is how you make the product on your best day—then force that behavior on your worst day.

“If MES is the bouncer at the door, the MBR is the guest list: who gets in, in what order, and under what limits—then leaves a trail proving it.”

TL;DR: The MBR is the master, version-controlled recipe that Quality approves and MES enforces. It defines materials, targets, tolerances, gravimetric controls, equipment and cleaning status, IPC and release testing via LIMS triggers, labels and verification, signatures with meaning, and explicit rework branches. Authored under Document Control and Change Control, executed with immutable audit trails per 21 CFR Part 11 and Annex 11, it is the operational source of truth for compliant, right-first-time manufacturing.

1) What an MBR Really Is—and What It Is Not

An MBR is the master template from which specific batch records are generated. It is scoped to include the approved bill of materials (BOM) with potency adjustments, unit-of-measure conversions, and rounding precision; the stepwise routing with exact sequencing, allowable concurrency, and mandatory line-clearance checkpoints; required equipment and tooling with fit-for-use criteria—calibration, cleaning validation, IQ/OQ/PQ, and environmental monitoring ranges; and definitive IPC and release testing plans that specify sample sizes, containers, preservation, and methods by version. It defines labels and print logic under Labeling Control, plus who signs what and when using role-based meaning of signature. Crucially, it also defines what to do when things go wrong: explicit rework and deviation branches, negative nets for removals, and instructions for hold times and quarantine. What it is not is a “guide” that leaves room for interpretation, local spreadsheets, or “tribal knowledge” to fill gaps. If a step can’t be enforced by the MES, the master is incomplete.

2) Master vs. Instance: MBR → eBMR Compilation

The practical distinction that trips people up is the difference between the approved master and the batch instance. The MBRDocument Control with version and effectivity dates, while the eBMR is compiled from that master for a specific lot, order, or campaign. Compilation inside the MES resolves parameterization—e.g., scale size based on batch size, directed picking rules from the eMMR, FEFO material selection, and which IPCs are hard stops. It binds device drivers for gravimetric steps, captures raw tare/gross/net readings, and checks instrument status in real time against calibration/cleaning records. The eBMR enforces scan-before-weigh to eliminate the “right weight, wrong item” failure mode, stamps every action with attributable user/time/device identifiers, and refuses to proceed when hard gates are unmet. That last clause matters: if your eBMR can be coerced by “supervisor override,” you don’t have a compliance system—you have wishful thinking with an audit trail.

3) Tolerances, Rounding, and Approach Rules—Designed for Enforcement

Most investigations around weighing, charging, and mixing trace back to sloppy tolerance design and hidden rounding. The MBR must state not just the nominal target but the acceptance window, the rounding precision used for display vs. acceptance, and any approach-from-below rules for potent actives with no positive deviation. Tare behavior must be defined and controlled so that operators cannot “walk the tare” to hide overfills. For materials dispensed via bench or analytical balances, require stability flags and dwell time before acceptance; for loss-in-weight feeders, define rate-of-change alarms and SPC alert/action limits. Expand tolerance rationale in the master—linking to clinical significance, process capability, and historical CP/CPK—so that Quality can defend why the window is what it is. If rounding rules differ between the MES UI and the stored value, formalize that behavior and show examples inside the MBR annex; otherwise you will eventually accept a value that displays “in” while the raw is OOS, and the auditor will make you explain why your system lies to operators.

4) Materials, Potency, and Lot Selection—From BOM to Reality

Every material row should be treated as a risk statement, not just a quantity. The MBR needs to capture potency or assay adjustments, allowable container types, and whether splits, merges, or kit pre-weighs are permitted. It should reference Dynamic Lot Allocation logic and reservation rules so that WMS location control can produce candidates by FIFO/FEFO while respecting quarantine and component release status. Identity confirmation via barcode validation must be mandatory and bound to the reservation, not merely a free scan. For pre-weigh operations, the MBR should define how kits are labeled, how variance is reconciled, and how batch-to-bin traceability is maintained when kits are staged to multiple orders. These rules, when explicit, collapse the most expensive class of errors—using the wrong thing perfectly.

5) Equipment, Cleaning, and Environment—Fit-For-Use by Design

The MBR must declare equipment classes and enumerated alternatives, not just a friendly name. It should include criteria that define “fit for use”: calibration in tolerance and within interval, cleaning status verified and appropriate to the product family, and qualification current for the mode of operation. If dry contact parts differ from wet contact parts, spell it out. If shared assets require cross-contamination control limits or campaign sequencing, put those constraints in the master so the MES can gate execution. Environmental conditions—temperature, humidity, differential pressure, airflow—should be set with acceptable ranges and tie to EM sampling points and alert limits. The MBR should require photo evidence or attachment upload for line clearance at initial setup and after interruptions. None of this belongs in tribal notes; it belongs in the master because the master is what the system enforces and the auditor reads.

6) IPC and Release Testing—Planned, Triggered, and Hard-Gated

IPC and release testing only keep you safe if they stop the process when results are unacceptable. The MBR should specify which IPCs are advisory and which are hard gates that hold the route until acceptable results are posted. For each test, embed sample size, containers, preservation, and method/version so that the LIMS can generate work, capture results, and post Hold/Release disposition back to MES/WMS. Sampling from blends or bulk intermediates should define sampling locations or homogenization steps to avoid the illusion of control created by “grab samples” that don’t represent the lot. When the MBR names reportables and release specs up front—including units, rounding, and decision rules—the LIMS can produce a defensible CoA without spreadsheet gymnastics. Tie these decisions back to CPV so long-term drift is visible before it turns into deviations.

7) Labels, Serialization, and Verification—Preventing Quiet Chaos

Label mistakes create a special kind of pain: everything appears correct until the wrong barcode or claim ricochets through receiving, WIP, and shipping. The MBR must define label templates under Document Control with versioning, variable fields, and conditional content tied to batch parameters, equipment, or packaging size. It should mandate verification at print (checksum/readability) and at point-of-use (content and reservation cross-check) and specify when GS1 GTIN, serials, or EPCIS events are required. For reprints or relabels, force scan-back of the prior label as a prerequisite, and capture photos for audit trail context. If your label logic lives in Excel, accept that you are running a voluntary recall simulation every day.

8) Governance: Document Control, Change Control, and Validation Reality

Great masters don’t write themselves. Establish an authoring process with drafting, technical review, QA approval, and controlled release. Every change must pass through formal Change Control with risk assessment, impact to CSV, training plans, and staged effectivity so open batches finish on the prior version while new batches start on the updated one. Validation should follow GAMP 5—risk-based, evidence-driven, focused on fitness for intended use—while electronic records and signatures must comply with Part 11 and Annex 11. The MBR repository should implement retention/archival policies and keep immutable history. If a change isn’t worth documenting, it wasn’t worth doing; if it was worth doing but wasn’t documented, you didn’t do it.

9) Genealogy, Splits/Merges, and Rework—Tracing Every Gram

Real factories split, merge, and rework; paper fantasies don’t. The MBR must declare how intermediates can be split across batches, how partials are reconciled, and how merges are tracked so that genealogy can be reconstructed from finished goods back to dispenses and raw lots without gaps. Rework, if allowed, should be a branch with specified criteria, additional sampling, and quality approvals—not an operator’s improvisation. Any material removal must be recorded as a negative net against the same step with reason codes and verifier sign-off. Require label scan-back when re-identifying containers so traceability isn’t broken by a new sticker. When these rules are embedded in the master and enforced by the eBMR, investigations stop being archaeological digs and become structured reviews with tight scope.

10) Metrics That Prove the MBR Is Working

Measure Right-First-Time lots (no reweigh, no rework), the frequency and type of block events (identity mismatch, equipment out-of-status, IPC failures), approach error on accepted additions, label reprint and scan-back failure rates, deviation density by step and material, review/approval cycle time, and release lead time. Correlate these to shifts, training completion, and equipment classes to reveal where the master is too loose or too tight. Feed results into CPV and APR/PQR. If your metrics rely on hand-groomed spreadsheets instead of system data, you’ve already lost the argument that your controls are effective; fix the master so the data falls out of normal work.

11) Common Failure Modes—and How to Design Them Out

Ambiguous tolerances and hidden rounding: The UI shows a value as “in spec” while the raw stored value is out. Prevent this by defining rounding and acceptance logic in the MBR, harmonizing UI and storage precision, and capturing raw signals in the eBMR. Right weight, wrong item/lot: Operators hit the gravimetric target using a mis-picked container. Eliminate this with directed picking tied to reservations and mandatory barcode validation before the system will read from the scale. Out-of-status equipment: A balance that’s overdue for calibration or a vessel not cleaned for this product family gets used because “we needed to move.” Interlock steps on status and require photo evidence for line clearance. Shadow spreadsheets: Uncontrolled tools for nets, labels, or IPC trending undermine data integrity. Route devices into MES, integrate to LIMS, and prohibit uncontrolled exports. IPC that doesn’t gate: Tests are run but production continues regardless. Mark IPC and release tests as hard gates with Hold/Release logic wired to LIMS disposition. Rework without genealogy: Material is scooped out or re-added off the books. Force negative nets, reason codes, and verifier signatures; require relabel scan-backs and updated genealogy.

12) How This Fits with V5

V5 by SG Systems Global treats the MBR as the single source of truth that compiles into enforceable execution. In V5 MES, each master step becomes an eBMR step with built-in interlocks for barcode checks, scale signals (gross/tare/net and stability), asset status, and line-clearance evidence. For sampling and testing, V5 issues triggers to the laboratory via V5 QMS – LIMS Integration, where methods, specs, and reportables are managed. Disposition flows back as Hold/Release state updates that hard-gate the route—no compliant way to proceed without an acceptable result. Labels for pre-weigh kits, intermediates, and finished goods print from versioned templates governed by Document Control; label verification is enforced both at print and at use. All actions—operator entries, device readings, exceptions with photo attachments, and reason-for-change—are written as immutable audit trails consistent with Part 11 / Annex 11. The result is clean lot genealogy from finished goods back to specific dispenses and instrument states. Analytics within V5 surface approach error, block event rates by station or operator, and release lead time drivers, feeding APR/PQR and supplier scorecards without manual collation. In short: V5 compiles the master into something a line can actually run—and fail safely when it should.

FAQ

Q1. What’s the difference between an MBR and an eBMR?
The MBR is the approved, version-controlled master; the eBMR is the compiled, enforceable batch instance that binds users, devices, materials, and results with attributable data and audit trails.

Q2. Should IPCs always hard-gate production?
Not always, but the MBR must explicitly designate which IPCs are advisory and which are hard gates. For critical quality attributes or identity confirmation, gate it—wire to LIMS disposition and Hold/Release states.

Q3. Can we change the MBR mid-batch?
No. Use a controlled deviation with risk assessment for the in-process lot. Permanent changes go through Change Control and apply to future eBMRs.

Q4. How do we ensure equipment is fit for use?
Declare equipment classes and status criteria in the MBR; interlock steps on calibration, cleaning validation, and qualification. Block use automatically in MES when any status is out.

Q5. What proves MBR data integrity during inspection?
Immutable master version history, signed approvals, validated compilation, raw device readings with timestamps and user/device IDs, reason-for-change entries, and computer-generated audit trails compliant with Part 11 and Annex 11.


Related Reading
• Core Systems: MES – Manufacturing Execution System | LIMS – Laboratory Information Management System | ELN – Electronic Laboratory Notebook
• Execution & Records: eBMR | eMMR | Batch Weighing | Gravimetric Weighing
• Controls & Integrity: Barcode Validation | Control Limits (SPC) | Audit Trail (GxP) | Data Integrity
• Traceability & Release: Lot Traceability – End-to-End Genealogy | CoA | Finished Goods Release
• Compliance Foundations: 21 CFR Part 11 | Annex 11 | GAMP 5 | APR/PQR