Because Canadian cannabis is sold in opaque, sealed packaging, a shopper standing at a counter has almost nothing to compare two products on except the THC percentage and the price. You cannot open a package, and until 2026 you could not see through one either. That single constraint has made the THC number carry more weight than it was ever designed to carry, and it is why label accuracy has become the defining industry argument of 2026. This post sets out what the evidence on THC numbers actually shows, what the rules do and do not require, and what else is printed on a flower label that is more informative than most people realise.
Why the THC number ended up doing all the work
Every mandatory field on a Canadian cannabis label exists for a regulatory purpose. Only two of them are usable as comparison tools in a shop: the Total THC figure and, separately, the price.
Aroma is sealed in. Appearance is behind cardboard or opaque plastic for most formats. Terpene content is not a required declaration, so there is no printed figure to compare. Harvest date is not required. Cultivation method is not required. What remains is a number expressed to one decimal place, and a shopper reaching for the only continuous variable on the box is behaving entirely rationally.
The difficulty is that the number is a single value standing in for a naturally variable plant, and the evidence on how well it does that job is not reassuring.
What the evidence actually shows
Five separate pieces of evidence are worth laying out in order, because they measure different things and are often conflated.
One laboratory's documented variance
High North Laboratories in Woodbridge, Ontario, has stated it directly. Rick Moriarity, the laboratory's COO, told StratCann in January 2023: "We have tested some of the inflated numbers and have had as much as a 12% THC variance compared to the label claim."
The wording is worth reading carefully, because the source does not specify whether the twelve refers to percentage points of THC or a relative percentage of the label claim. It is a laboratory's own account of the largest gap it has measured, and we are reporting it as stated rather than converting it into a figure the source does not give.
The same reporting includes a second observation from Moriarity's work on sample homogeneity: roughly 25 percent variance in THC levels was found within a single branch, before any blending.
An eight-laboratory collaboration
In December 2023, eight Canadian analytical laboratories jointly tested 16 brands of 3.5 gram dried flower, four brands of non-infused pre-rolls and 15 brands of infused pre-rolls, in work led by Rob O'Brien of Supra Research and Development.
- Nine of the 16 dried flower samples measured more than 25 percent below their label claim.
- Two dried flower samples measured more than 40 percent below label.
- Only one of the four non-infused pre-roll brands exceeded a 25 percent variance.
- Only three infused pre-roll samples deviated beyond a 15 percent range.
O'Brien's conclusion pointed away from laboratory error: "if they do set reasonable limits, producers can be within 15% of label claim." His argument was that the gap arises from producers testing unprocessed flower rather than the finished product.
Variability inside a single batch
In July 2025, Pure Sunfarms published research examining THC variation within plant layers, between plant sections and across plants in the same batch. The headline finding was that only 30 to 41 percent of individual potency measurements taken from plant tops fell within the batch's own 99 percent confidence interval.
In plain terms: the majority of individual samples from a batch did not sit close to the number that batch was labelled with. The study's point was not that anyone cheated. It was that a single value cannot represent a plant that varies from the top of a cola to the bottom of the same stem.
A regulator's testing programme
Alberta Gaming, Liquor and Cannabis has run secondary testing on products sold in the province since 2020, sending samples to third-party laboratories quarterly on a random and complaint-driven basis. As of September 2025, it had found 35 non-compliant samples across the life of the programme.
| Reason for non-compliance | Samples | Breakdown |
|---|---|---|
| THC-related | 33 | 17 dried cannabis, 12 edibles, 3 extracts, 1 topical |
| Pesticide | 1 | Dried cannabis |
| Microbial | 1 | Dried cannabis |
Two things follow from that table. The overwhelming majority of non-compliance found by a provincial regulator running a multi-year testing programme was about the THC number rather than about contamination. And the absolute count is small, which is worth stating because the figure is sometimes quoted without its denominator.
Ontario's own response
The Ontario Cannabis Store launched a temporary THC testing programme on 4 January 2024, selecting high-THC cultivars for secondary testing and returning products with inconsistent results to the vendor for re-labelling. Several products have since been re-labelled with lower figures. OCS applied an acceptable variance of plus or minus 15 percent against the label claim, so a product labelled 20 percent THC would be treated as within range between roughly 17 and 23 percent.
An OCS spokesperson described the purpose as being "to gather information and evaluate industry reports of inaccurate THC label claims on legal cannabis products."
The gap in the rules that explains most of this
Here is the part that almost nobody states plainly, and it is the most useful single fact in this post.
The Cannabis Regulations set a variability limit for THC and CBD on cannabis extracts, cannabis topicals and edible cannabis. They do not set one for dried cannabis.
Section 97(1) requires that a cannabis extract or a cannabis topical "must not contain, in respect of any quantity or concentration of THC or CBD that is displayed on the label, less than 85% or more than 115% of that quantity or concentration." Section 97(2) does the same job for edible cannabis on a sliding scale.
| Class of cannabis | Permitted variability against the label | Source |
|---|---|---|
| Cannabis extracts and cannabis topicals | 85 to 115 percent of the displayed quantity or concentration | Cannabis Regulations s.97(1) |
| Edible cannabis, over 5 mg declared | 85 to 115 percent | Cannabis Regulations s.97(2) |
| Edible cannabis, 2 mg to 5 mg declared | 80 to 120 percent | Cannabis Regulations s.97(2) |
| Edible cannabis, up to 2 mg declared | 75 to 125 percent | Cannabis Regulations s.97(2) |
| Dried cannabis, including flower and pre-rolls | Not addressed in section 97 | Cannabis Regulations, s.97 as enacted |
That asymmetry explains a great deal. The plus or minus 15 percent figure that circulates in discussions of flower labelling is not a rule about flower. It is the extract and topical limit from section 97(1), which the Ontario Cannabis Store has adopted as the threshold for its own testing programme. A distributor's policy is a real constraint on what a producer can list in Ontario, but it is not the same thing as a federal requirement, and it does not apply outside that programme.
Health Canada published a Validating Analytical Methods in Cannabis Testing Guide on 15 September 2026. It builds on the international ICH Q2(R2) validation protocols and adds cannabis-specific performance criteria, including distinguishing closely related compounds such as delta-8 and delta-9 THC, identifying synthesis byproducts, establishing predictable instrument response and demonstrating repeatability. It is aimed at analytical laboratories. It addresses how a number is produced, not what variance from that number is permitted on a dried flower label.
What Total THC actually means
Before comparing any two numbers, it helps to know what is being counted.
A living cannabis plant contains very little delta-9 THC. It mostly contains THCA, an acidic precursor that converts to THC when heated. Total THC is a calculated figure that accounts for that conversion, so it represents the THC available after decarboxylation rather than the THC present in the sealed package.
The units differ by format, which catches people out more often than it should.
- Dried flower is generally expressed as a concentration, in milligrams per gram or as a percentage. 180 mg/g is 18 percent.
- Total per package is the concentration multiplied by the net weight. Health Canada's own example: a 30 gram package labelled 180 mg/g Total THC contains 5,400 mg.
- Pre-rolls may show a figure per gram, per unit, or both, which makes a pack of five and a single not directly comparable without doing the arithmetic.
We went through the per-gram and per-package distinction in more depth in what the numbers on the package mean.
Everything else that is printed, and what it tells you
A dried cannabis label in Canada carries fourteen mandatory elements. Most shoppers read two of them. Here is what the rest are for.
| Field | What it tells you | What it cannot tell you |
|---|---|---|
| Total THC and Total CBD | Concentration as measured at the time of testing, after accounting for acid conversion | Uniformity within the package, or how the product will behave for any individual |
| Net weight in grams | How much product you are buying, which is what price per gram is calculated from | How that weight is distributed if the package holds several units |
| Number of discrete units and net weight per unit | On a pre-roll pack, how many there are and how much is in each | How much of each unit is flower and how much is anything else |
| Packaging date | Approximately when the product was sealed, accurate to about seven days either way | When it was harvested. Harvest is nowhere on a Canadian label |
| Lot number | Which production run this unit belongs to, which is how a recall is traced | Any date |
| Class of cannabis | Which regulatory class this is, and therefore which rules apply to it | Anything about the material inside beyond the class |
| Recommended storage conditions | What the producer says the product needs. Producers do not all say the same thing | How the product was stored before it reached the shelf |
| Licence holder name, phone and email | Who made it, and a direct route to ask them a question | Where it was grown, unless the producer chooses to state it |
| Excise stamp | That duty was paid and the product came through the legal supply chain. Province-specific | Anything about quality |
| Standardized cannabis symbol | That the product contains more than 10 micrograms of THC per gram | How much more |
| Health warning message | A mandated warning in a yellow box. The specific message rotates | Anything product-specific |
| Intended use | For example, that the product is for inhalation | Anything beyond the format |
The fields that reward attention most are the ones that are comparable across products: packaging date, net weight per discrete unit, and licence holder. Two products can carry the same THC figure and differ by a year in packaging date, or by half a gram per pre-roll.
We take the label apart field by field, including the ones that appear only on some formats, in reading a cannabis label, line by line.
What is deliberately not on the label
Four absences matter more than anything present.
- Terpene content. Canadian labels are not required to declare it. Some producers print a terpene profile voluntarily, and when they do it is useful precisely because it is comparable. Most do not. We covered what these compounds are in terpenes explained.
- Harvest date. Not required anywhere. The packaging date is the only date, and the gap between the two can run to months.
- Cultivation method. Indoor, greenhouse and outdoor are not mandatory fields.
- Minor cannabinoids. Beyond THC and CBD, there is no general requirement to declare other cannabinoids, though producers may choose to.
The packaging date point is the one most often misread, and it has its own post: packaged on is not harvested on.
What you can now see through the package
The sealed-packaging constraint loosened slightly in 2026. Changes to the packaging rules permit a transparent window on some cannabis packages, which returns a small amount of visual information to the counter. What you can see through a window is limited and is not a substitute for a label, but it is the first structural change to the problem since legalisation.
Structure, colour, the presence of visible trichomes and the amount of leaf material relative to flower are all observable characteristics. They are physical descriptions of the material, not predictions about anything, and we are describing them as such. The rule changes themselves are covered in why your cannabis package has a window now.
How to compare two products without relying on the THC figure
A practical sequence, using only printed information.
- Convert both to price per gram. Divide the shelf price by the net weight. Pack sizes are the single largest driver of per-gram price, so a 3.5 gram and a 28 gram package are not comparable until you do this.
- Compare packaging dates on the actual units. Online listings show the product. The shelf holds specific units, and their dates vary.
- Check net weight per discrete unit on multi-packs. A five pack of 0.5 gram pre-rolls holds 2.5 grams. A three pack of 1 gram pre-rolls holds 3 grams.
- Read the class and the intended use. These determine which rules govern the product and what it is designed for.
- Note the licence holder. It is the only field that lets you track your own experience across products, and it is printed with a phone number and an email address.
- Treat the THC figure as one measured value with known variability, not as a ranking. Everything above indicates the number is a point estimate of something that varies within a batch and sometimes within a single branch.
Questions with printed answers
- What is the packaged-on date on this unit rather than on the listing?
- Is the Total THC figure given per gram or per package, and what is the net weight?
- On this multi-pack, what is the net weight per discrete unit?
- Does this producer print a terpene profile, and does this unit carry one?
- Does the package state a cultivation method?
- Is there another unit of this product in stock with a different lot number?
Staff can read any of those off the package with you. They cannot supply information the label does not carry, and they cannot give medical advice.
The short version
- Sealed opaque packaging leaves THC percentage and price as the only practical comparison tools, which is why the THC figure carries so much weight.
- An Ontario laboratory has reported measuring as much as a 12 percent THC variance against a label claim, and roughly 25 percent variance within a single branch.
- An eight-laboratory study in December 2023 found nine of 16 dried flower samples measuring more than 25 percent below label, and two more than 40 percent below.
- Research published in July 2025 found only 30 to 41 percent of individual measurements from plant tops fell within their own batch's 99 percent confidence interval.
- Alberta's regulator has found 35 non-compliant samples since 2020, 33 of them THC-related.
- The Cannabis Regulations set variability limits for extracts, topicals and edibles. Section 97 does not address dried cannabis.
- The plus or minus 15 percent figure people cite for flower is a distributor testing threshold, not a federal rule for that class.
- Health Canada published a Validating Analytical Methods in Cannabis Testing Guide on 15 September 2026, aimed at laboratories.
- Packaging date, net weight per discrete unit, class, storage conditions and licence holder are all printed, comparable and underused.
- Terpene content, harvest date and cultivation method are not required on any Canadian label.
Our address and hours are on the visit page, and what is in stock is listed under all products. Staff at 811 Dundas St W will go through a label with you field by field and tell you what is printed on the unit in your hand.
Sources
- Cannabis Regulations (SOR/2018-144), section 97, variability limits: laws-lois.justice.gc.ca
- Health Canada, Packaging and labelling guide for cannabis products: canada.ca
- Health Canada, How to read and understand a cannabis product label: canada.ca
- StratCann, More changes needed to address inflated THC levels, 13 January 2023: stratcann.com
- StratCann, Eight Canadian cannabis labs team up to test accuracy of dried flower, pre-rolls, infused pre-rolls, 18 December 2023: stratcann.com
- StratCann, New research uncovers the significant THC variability in dried cannabis flower, 14 July 2025: stratcann.com
- StratCann, Update on AGLC's cannabis testing program, 10 September 2025: stratcann.com
- StratCann, OCS: Decline in high-THC flower following launch of testing program, 10 April 2024: stratcann.com
- StratCann, Health Canada releases new analytical cannabis testing guide, 15 September 2026: stratcann.com
- StratCann, Is it time for THC ranges instead of exact percentages?, 7 July 2025: stratcann.com