CBD lab report explained

What the document is, and who paid for it

A certificate of analysis is produced by a testing laboratory that has been sent a sample and asked to measure it. It reports what was found in that sample, on that day, by a named method. That is the whole scope of the claim it makes, and understanding the limits of that claim is most of what separates a useful reader from a credulous one.

Here is something no other guide on this subject states plainly, so we will. The brand commissions the test and pays the laboratory. That does not make the document worthless, and it is how testing works across most of the consumer goods economy. It does mean the incentive to test a favourable sample exists, and it is the reason three unglamorous details carry more weight than the numbers do: whether the laboratory is genuinely independent of the brand, whether the certificate is tied to a batch code you can match against the pack, and whether the brand publishes certificates that show problems as readily as ones that do not.

The accreditation to look for is ISO 17025, which is the international standard for testing laboratory competence, sometimes accompanied by UKAS recognition in the UK. An unaccredited laboratory is not automatically producing bad work, but accreditation is the difference between a result you can rely on and a result you have to take on trust. A certificate that does not name the laboratory at all is not a certificate. It is a graphic.

One more limit worth naming early. A certificate describes a sample drawn from a batch. It does not describe your individual unit, and it cannot account for anything that happened after testing. It is strong evidence, not a guarantee, and anyone presenting it as a guarantee has overclaimed.

The step every other guide skips: units

This is the part where the certificate and the law fail to speak the same language, and it is why so many people come away from a lab report no better informed than they went in.

A UK CBD product must hold no more than 1mg of controlled cannabinoids in the container. That is a weight, and it is per container. A certificate almost never reports weights per container. It reports concentration, either as a percentage by weight or as milligrams per gram, because that is what the instrument measures. So the number on the document and the number in the law are in different units, and comparing them directly is meaningless.

The conversion is one multiplication. Milligrams per gram, multiplied by the weight of the container in grams, gives the total in milligrams. Oils weigh a little under a gram per millilitre, so a 10ml bottle is roughly ten grams and a 100ml bottle roughly a hundred. That means the ceiling expressed in the certificate's own units is about 0.1 mg/g for a 10ml container and about 0.01 mg/g for a 100ml one, ten times tighter for the larger pack.

Work the other direction and the point sharpens. A certificate reading 0.05 mg/g looks reassuringly small. In a 10ml bottle that is half a milligram, comfortably inside the limit. In a 50ml bottle the identical figure is two and a half milligrams, which is over. Nothing about the extract changed. The reader who only checked the concentration would have got it wrong twice, once in each direction.

If a certificate gives a percentage rather than milligrams per gram, multiply the percentage by ten to convert. One per cent by weight is ten milligrams in every gram.

The THC line is not the total THC

Open a cannabinoid profile and you will usually find more than one line with THC in the name. Delta-9 THC is the familiar one. Directly above or below it there is very often THCA, tetrahydrocannabinolic acid, and that line is frequently the larger of the two.

THCA is what the living plant actually produces. It converts into THC when heated, a reaction called decarboxylation, which is why the two are reported separately: the laboratory is measuring the sample as it arrived, not as it might end up. The conversion is not one for one, because the molecule sheds a carboxyl group and therefore some weight. The standard factor is about 0.877, the ratio of the two molecular weights.

So total THC is the delta-9 figure plus roughly 0.877 times the THCA figure. A certificate showing a delta-9 line of 0.02 mg/g and a THCA line of 0.09 mg/g has a total of about 0.099 mg/g, near enough five times the delta-9 number alone. In a 10ml container that is the difference between two tenths of a milligram and just under a milligram.

Not one of the guides we examined on this subject mentions THCA at all. Every one of them tells the reader to check the THC line. If your product is one that gets heated, which covers vape liquids and anything cooked with, the distinction is not academic.

THC is not the only controlled line

The second thing the standard advice gets wrong by omission is that the UK exempt-product test does not count THC. It counts controlled cannabinoids, plural, and the milligram ceiling applies to that total rather than to any single compound.

Cannabinol, which appears on almost every certificate as CBN, is itself controlled in the UK. It is not psychoactive in the way THC is and it forms naturally as an extract ages, which means it tends to rise over the shelf life of a product rather than falling. THCV is in the same position. An extract that was comfortably compliant when it was made can drift as it sits, and the compound doing the drifting is one that most guides never mention.

In practice this matters most for full spectrum products and for anything that has been in a warehouse a while. It is also a reason certificate age is a substantive question rather than a bureaucratic one: a document from eighteen months ago is not describing the same cannabinoid profile the product has today.

None of this makes CBN sinister. It makes it countable, which is the only claim being made here.

What ND actually promises

Certificates report absent compounds as ND, meaning not detected, or as below the limit of quantification. Both phrases are frequently read as zero, and neither means it.

Not detected means not detected by that instrument, using that method, at that laboratory's threshold. The threshold is the limit of detection, and it varies between laboratories by more than most readers would guess. A sensitive method might detect down to a few parts per million. A blunter one might not see anything below a tenth of a per cent, and would report a genuinely significant quantity as ND with a straight face.

This is why the useful question is never whether a product is THC-free. It is what limit of detection the certificate was run at, and whether that figure is printed on the document. A certificate that states its detection limit is telling you how much confidence its ND lines deserve. One that does not is asking you to assume.

One competitor guide does make this point well, so it is not a secret. It is simply absent from the other five we examined, all of which describe broad spectrum and isolate products as containing zero per cent THC without qualification.

The panels beyond potency

Cannabinoid content is the section everyone reads. It is also the simplest test to run, and a great many products described as lab-tested have had nothing else done to them.

A full testing programme covers four further panels. Heavy metals matter because hemp is a bioaccumulator that draws contaminants out of the soil it grows in, so lead, cadmium, mercury and arsenic are the standard four. Pesticide residues matter for the same reason. Residual solvents matter where extraction used one, which covers most ethanol and hydrocarbon processes, and a supercritical carbon dioxide extraction has less to answer for here. Microbiological testing covers yeast, mould and bacterial contamination, and matters most for anything botanical that has been dried rather than distilled.

The honest position is that not every panel is necessary for every product, and demanding all four on an isolate powder is not sophistication. What is not acceptable is a report that quietly omits panels while the marketing implies comprehensive testing. Read the scope section, not only the results, and notice what was never run.

Terpene profiles turn up on some certificates too. They are interesting, they say something real about a full or broad spectrum extract, and they have no bearing on compliance whatsoever. Do not let a long terpene table stand in for a missing heavy metals panel.

A walkthrough you can run in three minutes

Put the whole thing in order and it becomes a short sequence. It is worth running once properly on a product you already own, because after that it takes about as long as reading a nutrition label.

Match the batch code on the certificate to the code on the pack, and check the report date. Confirm the laboratory is named, is not the brand, and states an accreditation. Read the scope to see which panels were actually run. Then go to the cannabinoid profile and note four lines: delta-9 THC, THCA, CBN and the CBD figure you are paying attention to for a different reason.

Convert. Add roughly 0.877 times the THCA figure to the delta-9 figure, add the CBN figure, and multiply the result by the container weight in grams. Compare the answer to 1mg. If the certificate reports a percentage rather than milligrams per gram, multiply by ten first.

Finally, sanity check the CBD figure against the label. A 1,000mg product in a 10ml bottle should report around 100 mg/g. If the certificate says substantially less, the product is underdosed relative to its own packaging, which is a different failure from a compliance one and is considerably more common.

If any step cannot be completed because a figure is missing, that absence is the finding. Our own policy on which documents we hold and how to request the one covering your order is on the lab reports page, and we would rather you ran this on us than took our word for it.

What is genuinely unsettled

Two things about this subject are less solid than a confident guide would suggest, and both are worth knowing before you rely too heavily on any certificate.

There is no mandatory certification regime for UK CBD. No law requires a brand to commission a certificate, publish one, or use an accredited laboratory, and there is no standard format, so two certificates for comparable products can differ in which compounds they report, which units they use and whether they state a detection limit at all. Everything described on this page is best practice observed in the market rather than a legal minimum, and the gap between the two is where most of the market's problems live.

Detection limits are also not standardised. Because ND is defined by the instrument and method rather than by regulation, two laboratories can look at the same material and report honestly different things. That is not fraud, it is measurement, and it is the strongest argument for caring which laboratory produced the document rather than only what it says.

Our own position, stated plainly: we are a stockist, the certificates we hold are the ones issued to the brands and passed on by our distributors, and the work of holding a current per-batch document for every full spectrum line is open rather than finished. That is the honest state of it in July 2026.