One in three THC claims is higher than every lab test of that strain

"Up to 30% THC." Says who?#

You've seen the number. Big, round, printed in bold on the seed pack or the product page. Then you grow it, or buy it, and it's a pleasant evening rather than a trip to orbit. Was it your grow? Your tolerance? Your imagination?

We wanted a better answer than a shrug. CannaLib holds 1,276,466 lab test results from state testing programmes and published lab sets, alongside the THC figures that seed banks, breeders and catalogues publish. So we took every strain where we have both a seller's claim and at least three flower lab tests, and put them side by side. Think of it as holding every "up to 30%" sticker up against the tape measure.

How the check works#

A claim is the highest THC figure a seed bank, breeder or catalogue states for a strain. A lab result is a flower test reported by a licensed lab for product sold under that same name. For each strain we took the median of all its claims, the median of all its lab tests, and the full lab range from lowest to highest. (The median is simply the middle result: line everything up from lowest to highest and it's the one in the centre, so a single freak jar can't drag it around.)

We kept the comparison fair to sellers. A strain only counted if it had at least three lab tests and a lab range reaching 5% THC or more, because lower results are usually delta-9 THC reported without THCA and can't be compared with a total-THC claim. CBD strains were left out. Claims above 35% were thrown out as not credible; there were 72 of them.

If a claim is higher than every lab test of that strain, it's a number nobody we can find has measured.

The ten biggest gaps#

Think dating-profile height. Shaving an inch is one thing; claiming a height that every tape measure on file says you never reached is another. Here are the strains where the claim and the lab disagree most. In all ten, the claim isn't just above the typical result. It's above the single strongest test on record.

Familiar Fruit is the extreme case: sold with claims around 34% THC, its lab median is 14.8% and its strongest test is 20.5%. Apple Gelato is claimed at 35% and has never tested above 20.3% in our records.

One caution before anyone gets angry. A strain name isn't a genome. Different plants get sold under the same name, and a seed bank's line may be stronger than the clone a dispensary tested. This measures the gap between what's written and what's been measured. It doesn't judge any single seller.

It depends who's writing the number#

The gap isn't spread evenly. We compared every source with at least 20 checkable strains. We don't name them here, because the point is how catalogues write their numbers, not a league table of who to blame.

One catalogue put 61% of its claims above every lab test (more than six in ten), with a median gap of six points. At the other end, one matched the lab median almost exactly and only 12% of its claims overshot. Same plants, same labs, very different habits.

Labs aren't perfect either#

It would be easy to treat lab numbers as gospel. We don't. CannaLib runs quality checks on every test before it counts toward a range, and the checks catch plenty.

That's why a CannaLib strain card never shows one hero number. It shows the median of several tests and the whole range, with every test linked to the lab and the record it came from.

We aren't the first to notice the gap. A 2023 study that bought retail flower in Colorado and had it independently tested found most products came in well under the THC printed on the label [1]. What our data adds is scale: thousands of strains, 13 catalogues big enough to compare, and every comparison open to inspection. Receipts included.

Why it matters#

THC isn't the whole story. Terpenes, dose, tolerance and setting all change how a strain feels. But if you pick a strain by its number, that number should be one somebody actually measured.