Section 123.02 of the Cannabis Regulations is the reason most Canadian cannabis labels say "no expiry date has been determined." The provision lets a licence holder print an expiry date only when it holds data proving the dried flower keeps 80% to 120% of its THC and CBD content for the whole stability period. Contaminants have to stay within limits for that period too, under the recommended storage conditions. Without that data, the label gets the disclaimer. Domestic buyers tolerate it. Export buyers don't.
Cannabis stability testing is the study that generates that data. You store packaged flower from several batches under controlled temperature and humidity and pull samples at fixed intervals. Each pull is tested for potency, moisture, water activity, microbial load and (for export) terpene content against the release specification. The study runs on the ICH Q1A(R2) framework that pharmaceutical regulators already recognise. It produces the two things a German or Australian importer needs before they can label a product: a justified shelf life and a storage statement.
This guide covers how the study is designed, what Health Canada requires before an expiry date can appear, what EU and Australian importers ask for, and the four places where LPs usually get the study wrong.
What cannabis stability testing measures and how the study is built
A stability study answers one question: for how long, and under what storage conditions, does this packaged batch still meet its specification? ICH Q1A(R2) sets the conditions, the batch count and the timepoints that regulators in Canada, the EU and Australia accept. The guideline was adopted in 2003 and is still the applicable text in late 2026. ICH released a consolidated replacement guideline (ICH Q1) for public consultation in April 2025, and its work plan targets Step 4 adoption in November 2026. Until each region adopts it, Q1A(R2) is the standard an importer will quote back to you.
Three storage conditions do the work. Long-term storage shows what actually happens over the claimed shelf life. Accelerated storage stresses the product to predict degradation and to catch packaging failures early. Intermediate storage is the fallback when accelerated results show a significant change.
| Condition | Temperature and humidity | Minimum duration | What it tells an importer |
|---|---|---|---|
| Long-term | 25°C ± 2°C / 60% RH ± 5% (or 30°C / 65% RH) | 12 months of data to support a claim, then continued to the full shelf life | The real shelf life under labelled storage |
| Intermediate | 30°C ± 2°C / 65% RH ± 5% | 6 months, run only if accelerated fails | Whether the product survives warmer warehouses |
| Accelerated | 40°C ± 2°C / 75% RH ± 5% | 6 months | How fast potency and moisture drift under stress |
Q1A(R2) asks for at least three primary batches, tested in the container closure system that will reach the market. Long-term samples are pulled every three months in the first year, every six months in the second, then annually. Accelerated samples are pulled at a minimum of three points, typically at zero, three and six months. The timepoints matter less than the sampling discipline behind them. Every pull must come from the same batch and the same pack, and go to the same ISO/IEC 17025 accredited laboratory on the same validated method. Otherwise the trend line you present is noise.
Dried flower degrades in predictable ways. THC oxidises to CBN over time, and the rate climbs with heat, light and oxygen. Residual THCA converts to THC when the flower warms, which can make a measured THC result rise before total THC falls. Monoterpenes such as myrcene and limonene evaporate first, which is why a 12-month-old batch can pass potency and still smell flat. Moisture migrates until the flower equilibrates with the headspace of the pack. A study that only tracks total THC misses three of those four mechanisms, and an experienced EU QP will notice.
What Health Canada requires before an expiry date can go on dried cannabis
Under section 123.02 of the Cannabis Regulations, the holder of the processing licence that manufactured the product must hold data establishing a stability period. During that period, packaged dried cannabis stored under its recommended conditions has to maintain not less than 80% and not more than 120% of its THC content and CBD content. Microbial and chemical contaminants have to stay within the limits referenced in subsection 93(3). Only then may the label carry an expiry date. If the data doesn't exist, section 123(1) requires the statement that no expiry date has been determined.
Two details catch LPs. First, the regulation ties the stability period to the product as packaged under the Regulations. A study run on bulk flower in a food-grade bag doesn't support an expiry date on a nitrogen-flushed export jar, and it doesn't support one on a 3.5 g retail container either. Second, subsection 123.02(3) requires you to retain the stability document for at least two years after the last sale or distribution of any lot carrying that expiry date. Health Canada inspectors ask for it, and so do importers performing a supplier audit (see our guide to what EU importers check in a supplier audit).
The 80% to 120% window sounds generous until you look at how it interacts with potency variance. A batch released at 30.3% THC has a floor of 24.2% before the expiry claim fails. If your initial result already carries sampling and method variance of several percentage points, measurement error eats the window before degradation does. That is why the sampling plan behind the study matters as much as the storage chamber. Tight initial data buys you shelf life.
Health Canada doesn't prescribe ICH conditions for cannabis. It prescribes the outcome. In practice, LPs that intend to export run the study to Q1A(R2) anyway, because the importer's regulator will. Running it twice to two standards is the expensive option.
What EU and Australian importers expect in a stability data package
In Germany, imported cannabis flower is a medicinal product, and its shelf life is not optional. The EU-GMP framework in EudraLex Volume 4 requires the manufacturer to operate an ongoing stability programme (Chapter 6) so that the product remains within specification for the shelf life on the label. The importer's Qualified Person has to be satisfied that the batch will meet its specification until expiry before QP batch release, and the Canadian LP is the only party that can supply that evidence. A quality agreement normally assigns stability responsibility explicitly. If that clause is unfamiliar, read our guide to what Canadian LPs sign in a quality agreement.
In Australia, the TGA applies TGO 93, which requires medicinal cannabis products to conform to the standard throughout their shelf life. The sponsor importing your flower needs stability data in the market pack to support the expiry date it declares. Australian importers routinely ask for the full study protocol as well as the summary table, because the TGA can request it.
| Market | Is a shelf life mandatory on the label? | Who relies on the data | Reference |
|---|---|---|---|
| Canada | No. Expiry date allowed only with stability data, otherwise "no expiry date has been determined" | Processing licence holder, Health Canada inspectors | Cannabis Regulations s. 123.02 |
| Germany and the EU | Yes. Medicinal product must carry a shelf life supported by data | Importer's QP at batch release, EU-GMP auditors | EudraLex Volume 4, Chapter 6 |
| Australia | Yes. Product must meet TGO 93 for its full shelf life | Sponsor, TGA on request | TGO 93 |
The difference in posture shows up in importer questionnaires. Canadian buyers ask whether you have an expiry date. EU and Australian buyers ask for the protocol, the chamber qualification records, the method validation and the raw data behind every timepoint. Germany's GMP washing crackdown in 2026 raised that bar further. A shelf life claimed on a foreign study with no ongoing programme now looks like exactly the gap regulators are hunting for. Build the stability package to the same standard as the flower itself.
Where LPs get stability testing wrong, and how to design the export study
Most LPs stall at packaging, not chemistry. The study runs in whatever container the QA team had on hand, and 14 months later the importer's pack is a different jar with a different liner and a different headspace. The data doesn't transfer, and the shelf life has to be rebuilt from zero. The second stall is water activity. Flower that equilibrates above the importer's limit at month six fails the study on moisture before potency ever moves. That is why our water activity and moisture limits guide belongs inside the protocol.
The third stall is method drift. A study that spans three years will outlive at least one HPLC column, one analyst and often one laboratory. If the method isn't validated and the lab isn't accredited, the importer's retest will diverge from your trend line and you'll be arguing about which number is real. We covered that argument in detail in why EU retesting produces different COA numbers. The fourth is terpenes. Potency can hold for 24 months while the monoterpene fraction halves, and an importer buying a named cultivar for its profile will notice.
The export stability package: 5 elements
- Protocol approved before day zero. Batches, pack, conditions, timepoints, methods and acceptance criteria are fixed in writing and signed by QA. A protocol written after the first pull is not a protocol.
- Three batches in the commercial export pack. Same container closure system, same liner, same fill weight and same gas flush the importer will receive.
- Full panel at every timepoint. Cannabinoid potency (THC, THCA, CBD, CBN), moisture, water activity, microbial counts and terpene profile, all run by an accredited lab on validated methods.
- Trend analysis instead of a pass/fail table. Plot each attribute against time and project the point where it crosses the limit. The shelf life is set from that projection with a margin, rather than from the last timepoint that happened to pass.
- Ongoing programme after launch. At least one batch per year goes into long-term storage so the shelf life stays supported as cultivation, drying and packaging change.
Here is the opinion we'll stand behind: an LP that ships export flower without at least 12 months of long-term data on the commercial pack is asking its importer to carry the regulatory risk, and the importer knows it. The study is slow and unglamorous. It is also the single most convincing document in a supplier qualification file.
AlphaLeaf grows indoor, hand-trimmed flower in Montreal under a Health Canada Standard Cultivation licence. Every batch is tested by an ISO/IEC 17025 accredited laboratory and tracked through full batch traceability. Our export authorisation under the Cannabis Act is backed by the documentation, stability records and testing data that EU-GMP importers and Australian sponsors ask for before a batch is released.
If you're an importer qualifying a Canadian supplier and you want to see a stability protocol before you see a price list, contact our team. The data is ready. Ask for it.
Frequently Asked Questions
What is cannabis stability testing?
Cannabis stability testing is a controlled study in which packaged dried flower from several batches is stored at defined temperature and humidity. Samples are tested at fixed intervals for potency, moisture, water activity, microbial load and terpenes. The results establish how long the product stays within specification under its recommended storage conditions, which is what supports a shelf life or expiry date on the label.
Does Health Canada require stability testing for dried cannabis?
Health Canada does not require a stability study for every product. Section 123.02 of the Cannabis Regulations does prohibit an expiry date on dried cannabis unless the processing licence holder has stability data showing the product keeps 80% to 120% of its THC and CBD content and stays within contaminant limits for the stated period. Without that data, the label must state that no expiry date has been determined.
What storage conditions are used in a cannabis stability study?
Most Canadian LPs follow ICH Q1A(R2). Long-term storage runs at 25°C ± 2°C and 60% RH ± 5%, intermediate at 30°C and 65% RH, and accelerated at 40°C and 75% RH. Long-term data covers at least 12 months before a shelf life is claimed, and accelerated studies run for six months.
How many batches are needed for a cannabis stability study?
ICH Q1A(R2) calls for a minimum of three primary batches, each stored in the same container closure system that will reach the market. For export, that means the actual commercial pack the importer receives rather than bulk storage bags.
Do EU and Australian importers require stability data from Canadian LPs?
Yes. In the EU, cannabis flower is a medicinal product and the importer's Qualified Person needs stability data to support the shelf life before batch release under EU-GMP Chapter 6. In Australia, TGO 93 requires the product to meet the standard for its entire shelf life, so the sponsor needs the LP's data to justify the expiry date it declares.
Why can a cannabis batch pass potency but still fail a stability study?
Potency is only one attribute. A batch can hold its THC content for 24 months while water activity drifts above the importer's limit, microbial counts rise, or the monoterpene fraction evaporates. Any attribute crossing its limit ends the supportable shelf life, which is why export studies test the full panel at every timepoint.
