CBN in France: legal status and 3 sleep trials

CBN stands for cannabinol, a molecule with the formula C₂₁H₂₆O₂ (310.4 g/mol). Its ring A, the one that carries the methyl group at position 9, is fully aromatic. Δ⁹-THC has the same skeleton, but in THC this ring keeps four more hydrogen atoms (C₂₁H₃₀O₂, 314.5 g/mol). CBN is therefore not "old THC": it is a different molecule, which appears when THC loses those hydrogens as it oxidises.

It is also the first cannabinoid ever isolated. Thomas Barlow Wood, W. T. Newton Spivey and Thomas Hill Easterfield extracted it in 1896 from charas, the resin of Indian hemp, then described it in detail in 1899 in "Cannabinol. Part I". Robert Cahn proposed its structure in 1932, and Roger Adams and his colleagues confirmed it in 1940 by synthesising it. A 2021 review in Frontiers in Pharmacology recalls the 1896 isolation and Adams's work.

CBN identity card

The identifiers come from the public PubChem database.

CBGA, the acid from which the plant then derives the other cannabinoids, is covered in our article on CBG.

The 2021 review raises a point that many product pages omit: the plant does not synthesise CBN. It comes from the breakdown of Δ⁹-THC, which ultraviolet light and heat convert bit by bit, even under good storage conditions. A fresh flower therefore holds very little of it. A flower or resin kept for a long time in air and light accumulates more as its THC falls.

CBN also has an acid form, CBNA, which the ANSM describes as its precursor. CBNA carries one extra carboxyl group; when it loses it as CO₂, it becomes CBN. Our article CBD or CBN puts figures on the THC lost by cannabis stored for several years, and our guide to the best-before date explains how oxidation tips the THC of a resin towards CBN.

In the laboratory, ageing is not needed. An Italian team showed in 2018 that iodine cleanly converts into CBN Δ⁹-THC, but also CBD and their acid forms, including from crude extracts of fibre hemp. The same team published in 2019 a one-pot total synthesis of CBN from citral and alkylresorcinols, molecules that are not cannabinoids. The 2021 review also lists semi-synthesis and full synthesis among the production routes for minor cannabinoids.

The name "CBN" on a label therefore says nothing about how the molecule was obtained. Only a certificate of analysis shows the THC that remains, and whether HHC, H4CBD or THCP are present, if the lab tested for them.

Few studies have given CBN to humans, and they are old or small. They describe a weak or absent effect when CBN is taken alone, and a slight strengthening of THC's effects when the two are combined.

In 1975, researchers gave five male volunteers, orally and one week apart, a placebo, 50 mg of CBN, 25 mg of THC and two combinations (12.5 mg of THC with 25 mg of CBN, then 25 mg of THC with 50 mg of CBN). THC raised the heart rate, CBN did not. On a questionnaire of 66 adjective pairs, the volunteers said they felt drugged, drunk, dizzy and drowsy under THC, but not under CBN. With the combinations, they felt more drugged, drunk, dizzy and drowsy than with THC alone, an increase the authors judge to be small. With five participants, the study gives an indication and does not allow a conclusion.

The 2026 insomnia trial, described below, qualifies this picture. At 300 mg, CBN raised the "feel effect" score one hour after the dose, and the "like effect" score the next day; the other scales, including "feel high" and "feel sedated", did not move. The authors note that this result contradicts most older studies, which described CBN as non-intoxicating, and recall that a 1973 intravenous study had already reported a weak psychoactive effect at high doses. Also at 300 mg, reaction time on the next-day vigilance test lengthened, and some driving-simulator measures, such as the distance to the vehicle ahead, changed, with no difference in lane keeping.

Receptor data point the same way. In a single 1997 study on cultured cells, CBN bound to CB1, the receptor behind the psychoactive effect, with an affinity constant of 211 nM, against 66.5 and 80.3 nM for Δ⁹-THC: its affinity was about three times weaker. At CB2, the gap was of the same order (126 nM against 32 nM). Reaching half the maximum effect at CB1 took 120 nM of CBN against 11 nM of THC. From one lab to another, the published values for CBN at human CB1 range from 13 to 1,100 nM in the ChEMBL database, which is why figures from a single experiment are the ones worth comparing.

For you, the practical risk of a product sold "with CBN" therefore lies less in the CBN than in what comes with it: residual THC, or a derivative sold under a similar name (see the legal status below).

The other effects attributed to CBN online, on pain, inflammation, appetite or bacteria, rest on work in rats, in mice or in the laboratory, on cells and bacteria, according to the 2021 review. These are not results obtained in humans.

Yes, for the molecule itself. The ANSM decision that classified a chemical family of cannabinoids as narcotics in 2024 excludes CBN by name, as it does its acid, CBNA. Its derivatives are not covered by that exclusion, and no food containing CBN is authorised.

The ANSM decision of 22 May 2024, published on 24 May and in force from 3 June 2024, adds to Annex IV of the order of 22 February 1990 "any substance derived from the benzo[c]chromene nucleus", whether its ring A is hydrogenated or not, wholly or partly, and whether or not it is substituted at the positions the text lists. The list ends with these words: "excluding cannabinol or CBN". The decision of 3 June 2024, published on 4 June, adds "cannabinolic acid or CBNA" to that exclusion, along with the THCA of compliant hemp. On our reading of the text, CBN meets the chemical definition of the family, and the text takes care to set it aside.

Keep it apart from the semi-synthetic molecules classified by the ANSM, the French medicines agency: HHC and two derivatives from 13 June 2023, then H4CBD, THCP and other derivatives from 3 June 2024. If a "CBN" product contained any of them, those decisions would apply to it. Nor is CBN a tetrahydrocannabinol, since its ring A is aromatic: on our reading of the text, the entry "Tetrahydrocannabinols, their esters, ethers, salts", moved to Annex III by a decision of 6 February 2026, does not cover it.

The exclusion names only CBN and CBNA. On our reading of the text, CBN derivatives that meet the generic definition have therefore been narcotics since 3 June 2024.

The abbreviation CBNO appears neither in the legal texts nor in PubChem. By analogy with HHCO, which the ANSM names as HHC acetate, and with THCPO, its "O" normally denotes an acetate, here cannabinol acetate. The three letters "CBN" in a product name therefore say nothing about its status.

A product containing CBN must also stay within the hemp framework. The order of 30 December 2021 authorises varieties listed in the EU or French catalogue whose Δ⁹-THC content is not above 0.30%, and applies that ceiling to extracts and to the products that contain them. A CBN-rich extract may come from degraded THC, which is why you need to check the remaining THC level on the certificate of analysis.

Products to swallow follow a different rule. On 20 May 2026, in a statement titled "Foods containing cannabidiol (CBD): illegal products that must be withdrawn from the market", the Ministry of Agriculture recalled that such foods are not authorised under Regulation (EU) 2015/2283 on novel foods. It announced checks extended to all foods that contain or advertise "CBD or any other cannabinoid", food supplements included. The Union list of authorised novel foods, in its version consolidated on 10 August 2026, contains no cannabinoid either. A CBN gummy, capsule or oil to swallow is therefore not authorised as a food, any more than its CBD equivalent. Our articles on the withdrawal of CBD gummies and on CBD regulation in 2026 set out this framework. Gummies that combine CBN and THC add a narcotics problem, as our article on CBN + THC gummies explains.

Status checked on 28 October 2026 against the ANSM consolidated list of narcotics, updated on 26 June 2026. The ANSM states there that the version of the 1990 order available on Légifrance does not include its decisions since February 2022: together with that order, it is those decisions, published on its website, that are legally binding.

We found three placebo-controlled trials that tested CBN alone on sleep. The two in the table missed their primary endpoint: the first was run by authors affiliated with two companies that sell cannabis products, the second by Australian researchers with university funding, with 20 patients and one night per dose. The third, run for the maker of the product tested and measured through online questionnaires, is described after them.

CBN was called the "sleep cannabinoid" before any data existed. In 2021, the review cited above still noted that the evidence for an effect of CBN on sleep was thin.

The first trial, published in Experimental and Clinical Psychopharmacology, compared a placebo, 20 mg of CBN alone and three CBN + CBD combinations (10, 20 or 100 mg of CBD). CBN alone lowered the number of awakenings and the overall sleep disturbance score, and adding CBD brought no improvement. The time taken to fall asleep and next-day fatigue did not change. Its nine authors are affiliated with Canopy Growth Corporation and Wana Brands, two companies that sell cannabis products.

The second trial, published in 2026 in the Journal of Sleep Research, used more rigorous measurement on a very small sample. At 300 mg, fifteen times the dose of the earlier trial, CBN shortened sleep onset by 7 minutes and slightly reduced arousals; at 30 mg, neither sleep onset nor night-time waking changed. The authors themselves note the limits: 20 participants, one night per dose and many outcomes tested. They conclude that their trial gives no direct evidence that CBN works against insomnia in routine care. The study was sponsored by the Woolcock Institute and funded by the Lambert Initiative at the University of Sydney, a philanthropically funded cannabinoid research centre; several authors declare consulting fees, including from an Australian medicinal cannabis industry association. Across all nights, the 20 participants reported 247 mild or moderate adverse events, 83 of them on placebo.

The third trial, published in 2024 in Pharmaceuticals, randomly assigned 1,020 American adults recruited online, who wanted to improve their sleep, to a placebo, melatonin or one of three doses of CBN, then followed them for five weeks. Everything was done remotely: sleep was measured only by a questionnaire filled in online, with no laboratory recording. The three CBN groups and the melatonin group reported less disturbed sleep than the placebo group, with no difference between them. The trial was run by employees of a research company under contract with the maker of the product tested, and about a quarter of participants completed no follow-up questionnaire.

On this basis, no effect of CBN on sleep is established. No health authority has recognised one, and its sale may not come with any health claim. Under Article L5111-1 of the French Public Health Code, a product presented as having curative properties is a medicine by presentation. For insomnia, the HAS, France's national health authority, sets out three steps, sleep hygiene first, then cognitive behavioural therapy, then a sleeping pill as a last resort, with the sleeping pill limited to four weeks, and CBN has no place at any of them. If your insomnia lasts, talk to a doctor. Our review of CBD and sleep applies the same reading to CBD.

With a CBN product, the first risk is whatever else it contains. In June 2025 the ANSM and ANSES, the French food safety agency, reported several hundred poisonings since 2024 linked to products sold as CBD. Synthetic cannabinoids or THC above 0.3% were usually the cause. The two agencies also note that CBD taken with some medicines can lower their effectiveness or add to their side effects.

We found no human trial on interactions between CBN and medicines, but its metabolism has been studied in the laboratory. In human liver microsomes, it is mainly broken down by two cytochrome P450 enzymes: CYP2C9, which turns it into 11-hydroxy-CBN, and CYP3A4, which turns it into 8-hydroxy-CBN. The authors of the 2026 trial point out that common medicines, such as some antidepressants, alter the activity of cytochrome P450 enzymes, which could increase exposure to either drug or next-day impairment. In mice, a 2025 study also showed that CBN slows the breakdown of melatonin by another enzyme, CYP1A2, and increases blood exposure to melatonin fourfold; this effect has not been checked in humans. If you take any treatment, ask a pharmacist for advice.

Pregnant or breastfeeding women and minors are excluded, as a precaution.

Does CBN show up in a drug test?

The roadside saliva test looks for THC, and CBN is not its target. The real risk remains the THC in a full-spectrum hemp extract: as MILDECA, the French interministerial drugs mission, points out, a sample can test positive for THC after a product that may legally be sold, and the law sets no threshold for the offence of driving after narcotic use. The penalties are set out in our article on CBD and driving.

Urine tests for cannabis, used in particular in medical settings, look for a THC metabolite, THC-COOH. A University of Utah team showed in 2020 that two of these tests also react to CBN. With the EMIT II Plus test, about five times more CBN than THC-COOH is needed to reach the positive cut-off, and the effect adds to that of THC; with the Microgenics MultiGent test, twenty times more is needed. According to the authors, these data can help decide on a confirmation test when a result does not match expectations; a method such as mass spectrometry identifies the THC metabolite itself. We found no equivalent study on roadside saliva kits.

A certificate of analysis should show four lines: CBN as a percentage or in mg/ml, Δ⁹-THC at 0.30% or less, THCA (which turns into THC when heated), and the absence of HHC, H4CBD and THCP if the lab tested for them. Check that the batch number matches the one on the bottle.

If the certificate shows both CBN and CBNA, total CBN is calculated as follows: CBN + 0.876 × CBNA. The coefficient comes from the molar masses, 310.4 ÷ 354.4 = 0.876: when it becomes CBN, CBNA loses about 12% of its mass as CO₂. On a label or a certificate, "CBNO" or "CBN-O" does not mean CBN: by analogy with HHCO, the abbreviation normally denotes an acetate, which falls under the narcotics rules on our reading of the text.

Each section is covered in our guide to the certificate of analysis, and our laboratory page lists the catalogue products.