On 1 October 2014, the revised Chapter 6 of EudraLex Volume 4 came into operation. It added a new section on technical transfer of testing methods and firmer rules for out-of-specification results. Twelve years later, it's the chapter that decides whether a Canadian cannabis batch is certified for the EU medical channel or held at the importer's warehouse.
EU-GMP Chapter 6 quality control requires a cannabis LP to run a QC function that is independent of production and owns specifications, sampling, testing and the quality decision on every batch. Methods must be validated or verified, results trended, OOS and out-of-trend results investigated under a written procedure, stability monitored on an ongoing basis, and any method moved to another lab formally transferred. If you export flower under EU-GMP, this is the file an inspector or Qualified Person opens first.
What EU-GMP Chapter 6 requires from a cannabis LP
Chapter 6 opens with a principle most Canadian LPs underestimate: the independence of Quality Control from Production is described as fundamental. Clause 6.1 says every manufacturing authorisation holder needs a QC department under a qualified, experienced person, with control laboratories and enough resources to do the work reliably.
QC isn't confined to the lab. The chapter expects it to be involved in every decision that could affect product quality, and finished product assessment has to consider production conditions, in-process results and a review of the batch documentation, not just the final COA.
Where Canadian sites run into trouble
Health Canada sets testing and quality expectations through Good Production Practices in Part 5 of the Cannabis Regulations, built around a Quality Assurance Person. That structure works domestically. Under EU-GMP, though, an inspector wants to see a named head of QC whose authority doesn't report through cultivation or processing.
The friction shows up in smaller facilities. The same manager plans harvests, reviews the lab results and signs off the batch. Nothing in that sequence breaks Canadian rules, but it reads as a conflict of interest to an EU auditor. If one person can approve their own batch, your QC isn't independent yet.
Laboratory practice, sampling and method validation
Clauses 6.5 to 6.14 cover the day-to-day lab, its documentation and sampling. Premises and equipment must meet the Chapter 3 requirements, and laboratory documentation follows the principles in Chapter 4. The QC department needs ready access to:
- Specifications for each cultivar and pack format
- Sampling, testing and recording procedures, including worksheets and lab notebooks
- Calibration, qualification and maintenance records for instruments
- A written procedure for investigating OOS and out-of-trend results
- Test reports or certificates of analysis
- Environmental monitoring data where required, and validation records for test methods
Any QC record tied to a batch is retained under the same rules as the batch record itself. Our guide to batch records under EU-GMP Chapter 4 covers retention periods in detail.
Sampling and method validation
Samples must represent the batch they come from, and each container is labelled with its contents, batch number and sampling date. For dried flower, that means a written plan for how many units you pull, from which parts of the lot, and who does it.
Clause 6.15 is the one that catches potency labs. Test methods must be validated, and a lab using a compendial method it didn't validate itself has to verify that the method is suitable in its hands. A cannabinoid method validated on mid-potency material and then run on a 30% THC cultivar without verification across that range is a finding waiting to happen.
| Chapter 6 area | What EU inspectors ask a cannabis LP | Common Canadian gap |
|---|---|---|
| Independence (6.1) | Who heads QC, and who do they report to? | QC reports through operations |
| Documentation (6.7) | Show the OOS and OOT procedure | OOS procedure exists, OOT missing |
| Method validation (6.15) | Validation or verification report per method | Range doesn't cover high-potency cultivars |
| Trending (6.16) | Trend charts for potency and moisture | Results stored, never trended |
| Ongoing stability | Current stability protocol and data | One study at launch, nothing since |
| Technical transfer (6.37-6.41) | Transfer protocol with the receiving lab | No protocol with importer lab |
Most of these gaps come from missing paperwork, not weak testing. Paperwork is cheaper to fix before an audit than after one.
OOS results, trending and ongoing stability
Clause 6.16 requires results to be recorded, trended and checked for consistency with each other. That single sentence turns QC from pass or fail into a running data set. A moisture result that sits inside specification but drifts upward across five consecutive batches is an out-of-trend signal, and Chapter 6 expects you to investigate it rather than file it.
Out-of-specification results follow the same written procedure. Good practice is to establish whether the cause sits in the lab or in the product before anyone decides on retesting. Our four-phase guide to cannabis OOS investigations walks through that sequence.
The ongoing stability program
Chapter 6 also requires an ongoing stability program after a product is on the market. The aim is to detect any stability issue across the shelf life in the pack you actually ship. For flower, that usually means cannabinoid content, moisture or water activity, microbial counts and appearance at set intervals.
Germany's Medizinal-Cannabisgesetz took effect on 1 April 2024 and moved medical cannabis out of the narcotics framework. Medical flower in Germany is now handled as a pharmaceutical supply question, and supplier qualification follows the same logic. Expect importers to ask for trend data and stability reports alongside the COA. An LP that can't produce them loses the tender to one that can.
Technical transfer of testing methods and contract labs
Clauses 6.37 to 6.41 were the headline addition in 2014. Before a test method moves from one lab to another, the originating lab and the receiving lab agree a transfer protocol. It covers the method's original validation status, the samples and materials to be tested, the acceptance criteria and how any deviation will be handled. The results and conclusions go into a transfer report.
This matters because EU importers often retest on arrival. When the Canadian COA says one thing and the EU lab reports another, the first question is whether the method was ever formally transferred. Our article on why EU retesting and COA numbers differ looks at the usual causes.
Contract labs and accreditation
Testing at an outside lab is allowed, but Chapter 6 says the reason should be stated in the QC records, and the arrangement falls under the EU-GMP Chapter 7 rules for outsourced activities. That means a written contract, a qualified lab and oversight you can show an inspector.
ISO/IEC 17025 accreditation helps, but it doesn't replace a transfer protocol. If your THC method has never been formally transferred to your importer's lab, you don't have one method. You have two numbers waiting to disagree.
AlphaLeaf is a Health Canada Licensed Producer of indoor-grown, hand-trimmed cannabis flower based in Montreal. We hold export authorisation under the Cannabis Act, and our ISO-certified batch testing and full traceability records are built to support the QC review that EU importers and Qualified Persons carry out before certification.
If you're qualifying a Canadian supplier and want to see how our QC file is structured, contact our team to request the documentation. Start with the stability data. It tells you the most.
Frequently Asked Questions
Does EU-GMP Chapter 6 apply to Canadian cannabis LPs?
In practice, yes. Chapter 6 is written for holders of an EU manufacturing authorisation, but EU importers and their Qualified Persons expect the exporting site's quality control to meet the same standard, and an EU GMP inspection of a Canadian site will audit QC against it.
Is Health Canada GPP testing enough to meet Chapter 6?
Not on its own. Good Production Practices under the Cannabis Regulations cover testing and quality assurance, but Chapter 6 adds independent QC, method validation or verification, result trending, out-of-trend investigation, an ongoing stability program and formal transfer of test methods.
What is an out-of-trend result in cannabis quality control?
An out-of-trend result sits inside specification but departs from the pattern of earlier batches, such as moisture creeping upward over several lots. Chapter 6 requires a written procedure for investigating these results, the same way OOS results are handled.
Can a cannabis LP use a contract lab for Chapter 6 testing?
Yes. Chapter 6 allows outside laboratories when the reason is recorded in the QC records, and the arrangement must follow Chapter 7, with a written contract and evidence that the lab is qualified and overseen.
What does technical transfer of a testing method involve?
The sending and receiving labs agree a protocol covering the method's validation status, the samples to be tested, acceptance criteria and how deviations are handled. Both labs test, the results are compared against the criteria, and the outcome is recorded in a transfer report.
Who makes the release decision for cannabis exported to the EU?
QC decides whether a batch conforms to its specifications at the manufacturing site. For the EU market, a Qualified Person at the importing manufacturer then certifies the batch before it can be released for supply.
