cbdlovers
EN

The UK has a number for this. No ranking on page one uses it

cbdlovers editorial team 22 min

We measured the first ten results for this search. Two of them are shops, and one of those shops has titled its own product page “The Best CBD Oil in the UK 2026”.

The rest are affiliate lists and one tested review from a national newspaper’s shopping section. Only one result on the entire page — a pharmacy chain’s health hub — is not attached to a sale.

So we are not going to publish an eleventh ranking. We are going to publish the ruler.

Because the UK, unlike anywhere else in the world, actually has a number for this. On 12 October 2023 two government committees set a provisional acceptable daily intake for CBD, and they converted it into drops themselves.

10 mg a day. Roughly four to five drops of a 5% oil.

Not one ranking on the first page uses it.

What the ten results actually are

It is worth being specific, because the shape of the page explains why the number is missing.

The shop that ranks itself

Two of the ten are retail sites whose own page titles claim the superlative. One of them carries “The Best CBD Oil in the UK 2026” as the title of its shop. That is not a ranking; it is a shelf with a headline.

The tested review

A national newspaper’s shopping section carries a genuinely tested piece, at just over four thousand words. It is the longest and best-researched result on the page — and that section of the paper is funded by affiliate links, which is disclosed and is not a secret.

The affiliate lists

Two more results are “top brands” listicles from lifestyle sites. Same model, less disclosure.

The one that isn’t selling a list

A pharmacy chain’s health hub, explaining what CBD oil is. No ranking, no products in the article.

And what none of them contains

The FSA number. Not the 10 mg, not the drop conversion, not the purity condition attached to it, and not the reason it doesn’t cover most of what they’re ranking.

The number the UK produced first

This is the centre of the page, so it is worth quoting precisely.

Who set it, and when

On 12 October 2023, the Advisory Committee on Novel Foods and Processes and the Committee on Toxicity published a joint position paper establishing a provisional acceptable daily intake for CBD in foods.

It came out of the FSA’s call for data to support novel food applications: the applicants themselves submitted new toxicological and human evidence, and a joint subgroup reviewed it.

The figure

0.15 mg per kilogram of body weight per day — that is 10 mg of CBD per day for a 70 kg healthy adult.

The regulator’s own conversion

And here is the part that should be on every product page in the country and is on none of them. The committees translated it themselves:

approximately equivalent to 4–5 drops of an oil-based supplement product containing 5% CBD of ≥98% purity

A government committee did the drop arithmetic for you. Nobody selling you drops repeats it.

What they said happens above it

The concerns named, if intake exceeds the provisional ADI, are somnolence — excessive tiredness — and adverse effects on the liver.

Why it says “provisional”

Because it may be refined as further scientific evidence becomes available. The subgroup also identified data gaps, particularly around reproductive and developmental toxicity and immunotoxicity.

And the sentence worth keeping

“This is the first time a provisional ADI has been established as a daily intake value for CBD in the UK or anywhere in the world.”

The UK went first. That is a real advantage for a British buyer, and it is being wasted.

The catch: it only covers isolate

This is the finding that changes what a “best CBD oil” list would even mean, and it is in the same document.

The three groups

The subgroup split CBD products into three:

Which group the number covers

Group A. Only Group A. The subgroup has completed its work on Group A products, and the 10 mg figure belongs to them.

What it says about the other two

That Group B and Group C are “more challenging and require further technical evaluation of the existing evidence and data gaps”, with further advice needed from the joint subgroup.

It adds that how CBD is absorbed across the gut in different food matrices may also need to be factored into product-specific assessments.

What almost every product in these ten results is

Full spectrum or broad spectrum. That is Group B or Group C.

So the only intake number the UK has does not cover the format that the entire first page is ranking. That is not a technicality — it is the single most useful thing a buyer could know, and it sits unquoted in a public document.

The public list: the instrument the UK has

Britain has one tool here that no EU country has, and it costs nothing to use.

What it is

The FSA maintains a public list of CBD products linked to novel food applications that have been validated. It is published, it is searchable, and it names products and brands.

What being on it means

That a dossier exists and reached the stage of validation. It does not mean the product has been authorised, and the FSA is explicit that the list is not an endorsement of safety.

What it does mean, practically

That someone submitted evidence. It is a filter, not a verdict — and a filter is more than any ranking on the first page gives you.

What it does not cover

The list concerns food products. Vapes are not food, and cosmetics are not food. If you are buying either of those, the list will not answer your question.

How to use it

Before you buy, check the exact product name and brand against the list. It takes a minute and it is the only check in this entire market that a buyer can complete on their own, for free, before spending money.

One milligram per container — and the two conditions nobody quotes

The other UK-specific number, and the one most often quoted at half length.

Where it comes from

Not from CBD law — there isn’t any. It comes from the definition of an “exempt product” in regulation 2 of the Misuse of Drugs Regulations 2001.

The three conditions, in full

An exempt product is a preparation or product where:

(a) it is not designed for administration of the controlled drug to a human being or animal; (b) the controlled drug is packaged such that it cannot be recovered by readily applicable means, or in a yield constituting a risk to health; and (c) no one component part contains more than one milligram of the controlled drug.

Which one everybody quotes

The third. “Under 1 mg per container” is the phrase you will see in shops, and it is accurate as far as it goes.

Which one nobody quotes

The first. The exemption requires that the product is not designed to administer the controlled drug. A trace of THC that is an unavoidable residue of a hemp extract is one thing; a product built around delivering it is another.

And why all three matter together

Because they are cumulative, not alternative. A product satisfies the definition when it meets all three. Any page that presents the 1 mg figure alone is presenting a third of a definition.

Why we are not publishing a ranking

We should say this plainly rather than let it be inferred.

The conflict, stated

This site makes money when someone buys hemp products. A ranking published by us would be exactly the thing we have just described in the ten results: a party with a commercial interest telling you which product wins.

What an honest ranking would need

Independent laboratory analysis of each product, purchased anonymously, batch by batch, with the results published in full including the ones that came back badly. Repeated, because a batch is not a brand.

That is expensive and slow, and nobody on the first page of this search has done it.

What we would rather do

Publish the criteria, in the open, and let them be applied to anyone — including to us, later. A ruler is more useful than a verdict, and it does not go stale when a formulation changes.

What that costs us

Traffic, probably. The people who came here for a numbered list will leave. That is a fair trade for not adding an eleventh list to a page that already has four.

The scorecard: what we would check, in order

Nine checks. The order matters — the early ones eliminate faster.

1. Is it on the public list

A minute, free, before buying. Product name and brand, exactly as written.

2. Is there a batch certificate

From an independent laboratory, with the batch number matching the bottle in your hand. A general certificate from two years ago proves the company once had something tested.

3. Milligrams per millilitre, not percentage

The percentage is a concentration. What you need is mg/ml, because that is where a serving comes from and where price per milligram comes from.

4. Total CBD, calculated properly

The plant makes cannabidiolic acid, not cannabidiol. When both forms appear:

Total CBD = CBD + (CBDA × 0.877)

The coefficient is the ratio of molecular masses: decarboxylating, the molecule loses carbon dioxide and therefore weight.

5. THC and THCA, listed separately

Without the acid form you cannot calculate the total, and the total is the figure that matters for the 1 mg condition and for a workplace test.

6. The contaminant panels

Heavy metals, pesticides, solvent residues, microbiology. Hemp is a bioaccumulator: it draws heavy metals out of the soil and holds them in its tissue, so effectively that it has been trialled for cleaning contaminated land.

7. Price per milligram

Total milligrams in the bottle, divided by the price. It is the only number that compares a 10 ml bottle at 10% with a 30 ml bottle at 5% — and it usually reverses the shelf impression.

8. The carrier oil

MCT, olive or hemp seed oil. It affects taste, shelf life, and how the product behaves with food. It is also the most common source of digestive upset in the first week.

9. What else is in it

Melatonin, curcumin, valerian, vitamins. As soon as a product contains several actives, any experience is about the mixture — and it carries the cannabidiol’s name.

The arithmetic that settles the “strength” question

One of the questions Google itself raises on this search is what the strongest legal CBD in the UK is. It has a precise answer, and it is bounded at both ends.

Percentage is not quantity

5%, 10%, 20% and 30% mean roughly 50, 100, 200 and 300 mg per millilitre. A drop is about 0.05 ml, so roughly 2.5, 5, 10 and 15 mg per drop.

Two people saying “ten drops” can differ sixfold.

The ceiling at the top

The exempt product definition caps controlled cannabinoids at 1 mg per container. That is a hard ceiling on the THC side, and it applies to the whole bottle, not per serving.

The floor from the other direction

The FSA’s provisional ADI is 10 mg a day — which the committees themselves put at four to five drops of a 5% oil.

So what “strongest” means here

That the strongest legal product is bounded by a controlled-drug ceiling on one side and by a public health intake figure on the other, and that a 30% oil delivers the ADI in less than one drop.

If a product’s own serving suggestion puts you well past 10 mg a day, that is worth noticing before the marketing does.

What the evidence says, area by area

Because “best for what” is the question underneath “best”.

Anxiety

The strongest area. A 2024 meta-analysis drew eight eligible papers from 1,550 screened, with 316 participants, and found a large effect: Hedges’ g = −0.92.

The 95% confidence interval runs from −1.80 to −0.04. That upper bound is close to nothing, and the authors ask for caution because of the small clinical sample. The studies concern diagnosed disorders.

Sleep

A systematic review gathered 34 studies. Only two examined people whose main problem was a sleep disorder. That is not “it doesn’t work” — it is “almost nobody has asked the question the way you are asking it”.

Pain

Two 2024 reviews disagree. One concludes that in the large majority of studies no benefit appeared against the comparison. The other finds reductions of 42–66% — but some of the included studies gave CBD together with THC, which is a different product.

The dose these findings sit at

An evaluation devoted specifically to low oral doses concludes that therapeutic benefits become clearly visible from 300 mg a day, and that below that threshold larger, more robust trials are needed.

And the collision

Three hundred milligrams where benefits become visible. Ten milligrams as the UK’s provisional daily intake.

Thirty times. We have no clever sentence that reconciles those figures, and we are suspicious of pages that do.

Who the committees say should avoid it

This list is from the same document as the 10 mg, and it is more specific than the European equivalent.

CBD-containing foods should be avoided, without the advice of a medical professional, by:

The second entry is the one that surprises people, and the EU statement does not name it separately.

On the last entry: cannabidiol inhibits certain liver enzymes — CYP3A4 and CYP2C19 among them — that metabolise a large share of commonly prescribed drugs. If a medicine’s leaflet warns about grapefruit, the same enzyme pathway is involved. A pharmacist settles it in two minutes with the list in front of them.

Format by format, and which question each one changes

The word “best” hides a format choice, and the format changes the question more than the brand does.

Oil

The only common format that can be held under the tongue, which bypasses part of the first pass through the liver. It is also the format the FSA’s drop conversion is written for.

Capsules

Precise going in, unpredictable arriving: swallowed only, no sublingual route, and a fatty meal multiplies what is absorbed. The number on the box is stable; what reaches the blood is not.

Gummies

A food, with sugar and flavourings, and the same swallowed-only route. Whatever the label says, the per-sweet quantity is what matters, not the pack total.

Vapes

Two separate warnings. First, the FSA’s food list does not cover them, so the one free check a UK buyer can do does not apply. Second, single-use vapes have been banned in the UK since 1 June 2025 under environmental powers — a rule about waste, not about cannabinoids, which catches this category anyway.

Cosmetics

A different regime entirely, and a different question: skin is a barrier, and cannabinoids are lipophilic. Serious transdermal research works with engineered patches and permeation enhancers, not with cosmetic emulsions.

The questions Google itself raises here

Four questions appear above the results on this search. Each has a real answer, and none of them is a brand name.

“Which brand of CBD oil is best?”

There is no answer to this that anyone on the first page is in a position to give, including us. What there is instead is the nine-point scorecard above, and a product either satisfies it or does not.

“Which strength CBD oil is best?”

Strength is a concentration, not a quantity. A higher percentage means fewer drops for the same milligrams — less oil and less taste, not more effect per milligram. What matters for the wallet is price per milligram; what matters for control is granularity.

Bounded at both ends, as set out above: 1 mg of controlled cannabinoids per container at the top, and a 10 mg provisional daily intake figure at the other. A 30% oil reaches the ADI in under a drop.

“Which CBD oil is best for pain relief?”

The two 2024 reviews disagree, and part of the disagreement is that some of the studies finding benefit administered THC alongside. That is not a brand question — it is a question about what was actually in the bottle in the trial.

What a batch certificate should contain

Since check number two decides most of the others, it is worth spelling out what a good one looks like.

The identity block

Laboratory name, accreditation, date of analysis, and the batch number — matching the bottle, not the product line.

The cannabinoid profile

CBD, CBDA, THC, THCA, CBN and CBG as separate rows, each with a figure and a unit. A single row saying “CBD” leaves out the half you need for the 0.877 calculation.

The limit of quantification

It should be stated. In analytical chemistry zero does not exist — there is a limit below which the method cannot measure. “Not detected” without that limit is a claim without a scale.

The four contaminant panels

Heavy metals, pesticides, residual solvents, microbiology. For an extracted product the solvents panel is the most relevant of the four, and the one most often absent.

And what will not be on it

Melatonin, vitamins, botanicals. Those are not cannabinoids and do not appear in a cannabinoid determination. That is the ingredients list’s job — two documents, two different questions.

What a genuinely good UK product would look like

We would rather set this out than imply it, because it is also the list we intend to be held to.

On the public list, by exact name

Not “the brand is on the list”. The product.

A batch certificate published, not on request

Linked from the product page, matching the batch on the bottle, with the full cannabinoid profile including the acid forms, and all four contaminant panels.

Milligrams per millilitre on the front

And the total in the bottle, and the price per milligram. All three, without arithmetic homework.

An honest position on the ADI

A serving suggestion that acknowledges the 10 mg figure exists, states where the product sits relative to it, and does not quietly recommend six times it.

An honest position on the group

If a product is full spectrum, saying so and saying that the UK’s intake figure was set for isolate and does not cover it. That sentence costs a sale and buys the only thing that matters here.

And no claim the register does not carry

There is no authorised health claim for CBD anywhere in Europe. A product that needs one to sell itself is telling you something about its evidence.

How the UK figure compares with the European one

Worth putting side by side, because the two numbers are close cousins and are often confused.

The European figure

In February 2026 the European Food Safety Authority set a provisional safe intake level of 0.0275 mg per kilogram of body weight per day — about 2 mg a day for a 70 kg adult — with an uncertainty factor of 400, while stating that the available evidence remains incomplete.

The UK figure

0.15 mg/kg, about 10 mg a day. Roughly five times higher.

Why they differ

Different committees, different dates, different weightings of the same evidence base — and the UK went first, in 2023, on data submitted by novel food applicants. Neither is a maximum safe dose in the sense of a cliff edge; both are provisional risk-assessment figures.

What they agree on

That the number applies to high-purity CBD, that it is provisional, and that vulnerable groups should be excluded. Both name pregnancy, breastfeeding, medication and children.

What that means for a British buyer

That if you read a European article giving 2 mg, and a British one giving 10 mg, neither is wrong. They are two regulators’ provisional readings of an unfinished evidence base, and the honest use of both is as an order of magnitude, not as a dosing instruction.

What “tested” means in a tested review

The best result on that first page is a genuinely tested piece, and it is worth being precise about what testing does and does not establish.

What a tester can establish

Taste, texture, how the pipette behaves, whether the bottle leaks, how quickly the order arrived, whether the certificate was easy to find. All real, all useful, all verifiable.

What a tester cannot establish

Whether the effect came from the cannabidiol. A reviewer trying six oils knows which is which, knows what each cost, and knows what each promises. There is no control arm and there cannot be one.

The size of that problem, measured

In a Canadian crossover study, 43 healthy adults received CBD-free hemp seed oil in both sessions. In the session where they were told it contained CBD, they reported greater sedation.

Zero milligrams, measurable effect.

Which is not an argument against testing

It is an argument for reading a tested review for the things testing measures well — the pipette, the packaging, the paperwork — and not for the sentence about how it made someone feel.

How to read this page

We apply to ourselves the criteria we ask of labels.

Who wrote it and with what interest

This site makes money when someone buys hemp products. A page that refuses to publish the ranking the reader came for, and then points out that the UK’s only intake figure does not cover most of what is sold, works against that immediate interest. That is a fact to weigh the text with, not a merit badge.

Compared with what

Every figure here has a stated reference point: the ADI for intake, the 1 mg definition for the controlled-drug side, placebo for the studies, and the 300 mg threshold for the literature.

At what quantity

We give study doses and regulatory figures, not recommendations. Ten and three hundred point in opposite directions, and it is honest to show both.

What we are not claiming

We have not laboratory-tested any product on that first page. Everything above is drawn from public documents and published studies, and every one of them is linked below.

Where the answers are

At the foot of the page, with direct links: the ACNFP and COT lay summary, the Misuse of Drugs Regulations, the FSA guidance, four PubMed papers and the European assessment report for the one authorised medicine.

An honest test you can run yourself

You cannot set up a controlled trial at home. You can do considerably better than an impression.

Write down the starting point first

For two weeks before buying anything, note one or two figures a day about whatever you want to observe. Without that there is nothing to compare against, and memory of how you were three weeks ago is notoriously unreliable.

Change one thing at a time

Starting the oil, an earlier bedtime and less coffee in the same week produces a result that says nothing about any of the three.

Note the meals as well

A fatty meal multiplies oral absorption. On a swallowed product it is the most important variable after the quantity itself, and the one nobody records.

Give it longer than a week

Cannabidiol clears slowly, and with regular use the level takes days to settle. The first few days are also the worst possible measurement window, because that is when expectation is strongest.

And if you buy isolate, you can go one better

An isolate dissolved in carrier oil has no smell and no taste. Two identical bottles prepared by somebody else — one with, one without, labelled A and B by them — is the closest thing to a blind test available in a kitchen. It is the only format that allows it.

In short

The number

10 mg a day — 0.15 mg/kg — set by the ACNFP and COT on 12 October 2023, the first provisional ADI for CBD anywhere in the world. The committees put it at 4–5 drops of a 5% oil themselves.

The catch

It covers Group A only: products using ≥98% pure CBD. Full spectrum and broad spectrum are Groups B and C, described as “more challenging” and not yet assessed. That is most of what the first page ranks.

The free check

The FSA public list. One minute, exact product name, before you spend anything. It is not an endorsement — it is a filter, and it is the only one a buyer can run alone.

The definition to quote in full

Exempt product has three conditions, not one. The 1 mg per container is the third; the first is that the product is not designed to administer the controlled drug.

And why there is no ranking here

Because every ranking on that page belongs to someone selling, ourselves included. What we can usefully publish is the ruler — and then be measured by it.

If a product ever satisfies all nine checks above, the scorecard will say so and you will be able to verify each line yourself. That is a slower way to earn a first place, and it is the only one that survives someone checking.