1Where three batches came from.
Three batches is a convention, not a US requirement. Asked whether CGMP requires three successful validation batches before release, FDA answers “No. Neither the CGMP regulations nor FDA policy specifies a minimum number of batches to validate a manufacturing process.” It says validation “cannot be reduced to so simplistic a formula as the completion of three successful full-scale batches”, and that the idea of three batches “became prevalent in part because of language used in past Agency guidance”. The manufacturer is expected to have “a sound rationale” for the number it chooses.¹
FDA says its process validation guidance now recommends a product lifecycle approach, with the emphasis on process design and development studies as well as reproducibility at scale.¹ That guidance, of January 2011, sets no batch count. The approach to process performance qualification (PPQ) “should be based on sound science and the manufacturer’s overall level of product and process understanding and demonstrable control”, the decision to begin commercial distribution should be supported by data from commercial-scale batches, and PPQ must be completed successfully before commercial distribution.²
2What each authority says.
| Source | Status | On the number of batches |
|---|---|---|
| 21 CFR 211, US drug CGMP | Regulation | No minimum number specified.¹ |
| FDA CGMP questions and answers, question 24 | FDA questions and answers, for cell and gene therapy products | No set minimum number of PPQ batches. A BLA should include PPQ data, but the amount can be adjusted when justified.¹ |
| FDA process validation guidance, 2011 | Guidance | No number. Sampling should give “sufficient statistical confidence of quality both within a batch and between batches”.² |
| EU GMP Annex 15, 2015, sections 5.19 and 5.20 | EU GMP guideline | Each manufacturer “must determine and justify the number of batches”. A minimum of three consecutive batches under routine conditions is “generally considered acceptable”; another number may be justified.³ |
| PIC/S GMP Guide, PE 009-18, Annex 15 | PIC/S GMP guide | The same text as EU Annex 15.⁴ |
| PIC/S PI 006-4, October 2026 | PIC/S guidance for inspectors and industry | The number of consecutive batches “should be scientifically justified”; the three-batch wording in Annex 15 is “viewed to be transitory”.⁵ |
| ICH Q7, section 12.50 | ICH guideline for active substances | Three consecutive successful production batches “should be used as a guide”, with more where the process is complex or slow to complete.⁶ FDA’s reading is that Q7 does not specify a number.¹ |
| EMA process validation guideline, 2016 | Guidance on the dossier | Where production-scale data must be submitted, for example for biological products, at least three production-scale batches unless justified; one or two can suffice with supporting pilot batches.⁷ |
| WHO TRS 992 Annex 3, now Appendix 7 of TRS 1019 Annex 3 | WHO guideline | Three batches were traditionally the normal number, but the number “should be justified and based on a risk assessment”.⁸ |
PI 006-4 explains its position in a footnote: its recommendation builds on the Annex 15 wording but is meant “to advance the application of science and QRM”, and it expects the next update of Annex 15 to reflect a risk-based approach.⁵
3How to justify the number.
The sources point to the same factors.
- Variability. Annex 15 asks for enough batches to establish the normal range of variation and trends.³ PI 006-4 says increased variability or risk “should normally result in an increased number of batches” and more intensive sampling.⁵ WHO lists variability of results from process design, variability of materials, product history and where the product is being transferred from.⁸
- Complexity. ICH Q7 ties the number to the complexity of the process or the size of the change,⁶ and EMA to the variability and complexity of the process and product.⁷
- Prior knowledge. Annex 15 allows another number where standard methods of manufacture are used or similar products or processes already run at the site.³ FDA notes that “previous credible experience with sufficiently similar products and processes can also be helpful”.² PI 006-4 accepts that understanding intra- and inter-batch variability at full scale can support fewer batches, but not where data are less extensive.⁵
- Statistics. FDA “strongly” recommends objective measures, such as statistical metrics, wherever feasible, and a sampling plan whose confidence level reflects the risk of each attribute.²
- A defined end point. WHO asks manufacturers to define the stage at which the process is considered validated and the basis for that decision, with a justification for the number of batches.⁸
Write the rationale into the protocol. ICH Q7 expects the validation protocol to state the number of process runs,⁶ and batches made for validation should normally be the same size as commercial batches.³
4What happens after the validation batches.
The batch count is not the end of validation. Annex 15 says an initial exercise with three batches “may need to be supplemented with further data obtained from subsequent batches as part of an on-going process verification exercise”.³ PI 006-4 expects any weakness in the validation data to be compensated by additional testing during ongoing process verification.⁵ FDA recommends keeping PPQ levels of sampling and monitoring into continued process verification until there are enough data to estimate variability.²
5How it runs in Seal.
In Seal’s process validation blueprint, the PPQ protocol carries its scientific sampling rationale, acceptance criteria and planned analysis, and the campaign design needs its own justification. Execution, laboratory results and protocol exceptions are connected records, so review starts from open exceptions and evidence coverage rather than a batch count, and the reviewed controls and residual risks carry into continued process verification.
For cell and gene therapy, where processes change between early studies and routine production, each change becomes the next version of the workflow, assessed for its intended use and released under change control, while lots already made keep the basis they were made on.
References
- 1Questions and Answers on Current Good Manufacturing Practice Regulations: Production and Process Controls, questions 5 and 24, FDA, 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) - 2Process Validation: General Principles and Practices, Guidance for Industry, FDA, January 2011, sections II.B, IV.C.2, IV.C.3 and IV.D. https:/
/ www. fda. gov/ media/ 71021/ download (read 6 November 2026) - 3EudraLex Volume 4, Annex 15: Qualification and Validation, European Commission, in operation from 1 October 2015, sections 5.8 and 5.18 to 5.20. https:/
/ health. ec. europa. eu/ system/ files/ 2016- 11/ 2015- 10_ annex15_ 0. pdf (read 6 November 2026) - 4PIC/S Guide to Good Manufacturing Practice for Medicinal Products, Annexes, PE 009-18, Annex 15, PIC/S, 24 September 2026. https:/
/ picscheme. org/ docview/ 11333 (read 6 November 2026) - 5PI 006-4, Recommendations on Qualification and Validation, PIC/S, 1 October 2026, sections 5.2.6 and 5.5.2, with footnote 3. https:/
/ picscheme. org/ docview/ 11277 (read 6 November 2026) - 6ICH Q7, Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients, ICH, 10 November 2000, sections 12.21 and 12.50. https:/
/ database. ich. org/ sites/ default/ files/ Q7%20Guideline. pdf (read 6 November 2026) - 7Guideline on process validation for finished products: information and data to be provided in regulatory submissions, EMA/
CHMP/ CVMP/ QWP/ BWP/ 70278/ 2012- Rev1,Corr. 1, EMA, 21 November 2016, section 5.1. https:/ / www. ema. europa. eu/ en/ documents/ scientific- guideline/ guideline- process- validation- finished- products- information- and- data- be- provided- regulatory- submissions- revision- 1_ en. pdf (read 6 November 2026) - 8WHO Technical Report Series No. 1019, Annex 3, Good manufacturing practices: guidelines on validation, Appendix 7, Non-sterile process validation, as published in TRS 992, Annex 3, 2015, section 5. https:/
/ www. who. int/ docs/ default- source/ medicines/ norms- and- standards/ guidelines/ production/ trs1019- annex3- gmp- validation. pdf (read 6 November 2026)
AQuestions and answers
How many PPQ batches does a cell or gene therapy need?
FDA sets no minimum. A BLA submission is expected to include PPQ data, but the amount can be adjusted when justified, and where production is limited the number can rest on the complexity of the process, the controls in place and the manufacturer’s understanding of the product and process. Characterisation studies support that case, and so can experience with similar products. FDA encourages sponsors to discuss their plans with CBER.
Can we validate with fewer than three batches in the EU?
Yes, with a justification. Annex 15 allows a different number where standard manufacturing methods are used or similar products already run at the site, and EMA accepts one or two production-scale batches in a dossier when pilot-scale batches and a justification support them. PIC/S accepts fewer batches where variability at full scale is already well understood.
Do PPQ batches have to be consecutive?
Where a guideline names a number, it names consecutive batches: Annex 15 speaks of three consecutive batches under routine conditions and ICH Q7 of three consecutive successful production batches. PIC/S asks for the number of consecutive batches to be justified.
Do PPQ batches have to be at commercial scale?
Normally, yes. Annex 15 expects validation batches to be the same size as intended commercial batches unless another size is justified, and FDA says the decision to start commercial distribution should be supported by data from commercial-scale batches.