Skip to main content

Manufacturing and batch records

How do you calculate batch yield and reconcile materials and labels, and when must a discrepancy be investigated?

Percentage of theoretical yield is actual yield divided by theoretical yield, times 100, at each appropriate phase of manufacture. 21 CFR 211.103 requires it to be calculated and independently verified; the master record sets the maximum and minimum percentages beyond which 21 CFR 211.192 requires an investigation. Labelling issued, used and returned is reconciled, and discrepancies outside narrow preset limits are investigated. EU GMP asks for significant yield deviations to be investigated and reconciliation discrepancies accounted for before release.

Illustration of a seal beside a production line of sealed vials.

1How yield is defined and calculated.

FDA defines three quantities. Theoretical yield is what “would be produced at any appropriate phase” based on the components used, “in the absence of any loss or error in actual production”. Actual yield is what is actually produced at that phase. Percentage of theoretical yield is the ratio of the two at the same phase, “stated as a percentage”.¹

Percentage of theoretical yield = actual yield ÷ theoretical yield × 100

If a compression step charged with enough granulate for 100,000 tablets produces 97,400 tablets, the percentage of theoretical yield at that phase is 97.4%.

Under 21 CFR 211.103, actual yields and percentages of theoretical yield “shall be determined at the conclusion of each appropriate phase” of manufacturing, processing, packaging or holding. One person calculates and a second independently verifies; if automated equipment calculates the yield, one person verifies it.² The batch record states the actual yield and the percentage of theoretical yield at those phases, under 211.188(b)(7).² EU GMP asks the batch processing record to show “the product yield obtained at different and pertinent stages of manufacture”.³

For active substances, ICH Q7 asks for actual yields to be compared with expected yields at designated steps, defining expected yield as the quantity or percentage of theoretical yield anticipated at a phase “based on previous laboratory, pilot scale, or manufacturing data”.⁴

2Where the limits come from.

Part 211 does not set a percentage. It requires each master production and control record to include “a statement of theoretical yield, including the maximum and minimum percentages of theoretical yield beyond which investigation according to § 211.192 is required”.² ICH Q7 likewise asks for expected yields “with appropriate ranges” established from earlier data.⁴

FDA’s CGMP questions and answers address processes with high but consistent losses. Asked about processes such as transdermal patches with high in-process reject rates, FDA said lower yields are not of themselves a concern, but “validated and well-controlled processes will achieve fairly consistent waste amounts batch-to-batch”, and waste above normal operating rates may need evaluating for its cause and its consequences for product quality.⁵

3How materials and labels are reconciled.

A reconciliation compares what was issued to a batch with what can be accounted for. Part 211 requires, for each component, an inventory record with “a reconciliation of the use of each lot”.² EU GMP Chapter 5 asks for “checks on yields, and reconciliation of quantities” to ensure there are no discrepancies outside acceptable limits.⁶ For packaging, the EU batch packaging record holds the quantities of printed packaging materials and bulk product “issued, used, destroyed or returned to stock and the quantities of obtained product, in order to provide for an adequate reconciliation”.³

Part 211 requires “strict control” over labelling issued for a batch. Section 211.125(c) requires procedures to “reconcile the quantities of labeling issued, used, and returned”, and evaluation of discrepancies between the quantity of finished product and the labelling issued “when such discrepancies are outside narrow preset limits based on historical operating data”. Those discrepancies are investigated under 211.192. Reconciliation is waived for cut or roll labelling only when electronic or electromechanical equipment examines 100 percent of units for correct labelling. Excess labelling bearing lot or control numbers is destroyed.² In the EU, unused batch-coded packaging materials are destroyed at the end of packaging and the destruction recorded.⁶

CheckUnited StatesEuropean Union
Yield at each phase211.103, recorded under 211.188(b)(7)Chapter 4, 4.20(g)
Yield limits211.186(b)(7), in the master recordChapter 5, 5.44: significant deviation from expected yield
Component reconciliation211.184(c), for each lotChapter 5, 5.8
Labelling and packaging211.125(c), narrow preset limitsChapter 4, 4.21(h); Chapter 5, 5.61
Investigation211.192, whether or not the batch has been distributedChapter 5, 5.44; 5.61, before release

4When a discrepancy must be investigated.

Under 211.192, the quality control unit reviews and approves all production and control records, including packaging and labelling, before a batch is released. “Any unexplained discrepancy (including a percentage of theoretical yield exceeding the maximum or minimum percentages established in master production and control records)” must be thoroughly investigated, “whether or not the batch has already been distributed”. The investigation extends to other batches of the same product and to other products that may be associated with the discrepancy, and its written record includes the conclusions and follow-up.²

The wording covers both ends of the range: a yield above the maximum needs investigating as much as one below the minimum.

EU GMP sets the same expectation in its own terms. “Any significant deviation from the expected yield should be recorded and investigated”, and “any significant or unusual discrepancy” in reconciling bulk product, printed packaging materials and units produced “should be investigated and satisfactorily accounted for before release”.⁶ For active substances, ICH Q7 asks for deviations in yield associated with critical process steps to be investigated for their impact on the quality of affected batches.⁴

5How it runs in Seal.

In Seal, a batch material reconciliation compares the issued quantity with evidenced consumption, returns, samples and waste on the same quantity basis, and keeps an unexplained difference visible until evidence resolves it, as the worked example shows. Neil, Seal’s AI agent, can trace a material-balance discrepancy across ERP transactions, batch records and warehouse receipts into a source-linked report, and reconcile packaging components and labels so that unexplained labels remain visible before review. Neil does not create a loss entry or raise a reject count to make a balance close; your reviewers assess the discrepancy.

When yield falls, Neil can compare the runs with their parameters, timestamps and yields and draft an investigation with linked evidence and next checks. Your process owner assesses the explanations and decides what to test.

References

  1. 121 CFR 210.3, Definitions, eCFR. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-C/part-210/section-210.3 (read 6 November 2026)
  2. 221 CFR Part 211, Current good manufacturing practice for finished pharmaceuticals, including 211.103, 211.125, 211.184, 211.186, 211.188 and 211.192, eCFR. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-C/part-211 (read 6 November 2026)
  3. 3EudraLex Volume 4, Chapter 4: Documentation, European Commission, in operation 30 June 2011. https://health.ec.europa.eu/system/files/2016-11/chapter4_01-2011_en_0.pdf (read 6 November 2026)
  4. 4ICH Q7, Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients, ICH, 10 November 2000. https://database.ich.org/sites/default/files/Q7%20Guideline.pdf (read 6 November 2026)
  5. 5Questions and Answers on Current Good Manufacturing Practice Requirements: Production and Process Controls, question 4, FDA, page content current as of 5 May 2026. https://www.fda.gov/drugs/guidances-drugs/questions-and-answers-current-good-manufacturing-practice-regulations-production-and-process (read 6 November 2026)
  6. 6EudraLex Volume 4, Part I, Chapter 5: Production, European Commission, 13 August 2014, in operation 1 March 2015. https://health.ec.europa.eu/document/download/4a1fdb4f-6f6f-49c4-b264-8056e5bbe078_en?filename=chapter_5.pdf (read 6 November 2026)

AQuestions and answers

What is the formula for percentage of theoretical yield?

Divide the actual yield at a phase of manufacture by the theoretical yield at the same phase, and multiply by 100. Theoretical yield is what the components charged would produce with no loss or error. Both figures must use the same units and the same phase.

Is a yield above the maximum a problem?

Yes. Part 211 treats a percentage of theoretical yield above the maximum in the master record the same as one below the minimum: both are discrepancies that 211.192 requires to be thoroughly investigated, whether or not the batch has already been distributed.

What is the difference between theoretical, expected and actual yield?

Theoretical yield is the quantity the materials would give with no loss. Actual yield is what the phase produced. Expected yield, a term ICH Q7 uses for active substances, is the quantity or percentage you anticipate from earlier laboratory, pilot or manufacturing data, with a range around it.

Who calculates and checks the yield?

Under 21 CFR 211.103, one person calculates the yield and a second person independently verifies it. If automated equipment performs the calculation, one person independently verifies it.

Do cut labels have to be reconciled?

Yes, unless every unit is examined for correct labelling by electronic or electromechanical equipment during or after finishing. Without that 100 percent examination, cut and roll labels are reconciled like other labelling, against narrow preset limits based on historical data.

Does a yield investigation have to look at other batches?

Under 21 CFR 211.192, yes. The investigation extends to other batches of the same product and to other products that may be associated with the discrepancy, and it applies even if the batch has already been distributed.

See your process in Seal.

Bring a procedure or a recurring problem. See how your team configures it with Neil and improves the next run.

Book a demo