Cannabidiol

Cannabidiol at a Glance

Cannabidiol (CBD) is the major non-intoxicating cannabinoid of cannabis and hemp, and one of very few plant compounds that is simultaneously an approved prescription medicine and a mass-market supplement.

Its pharmacology is the reason it does not get you high: CBD works largely around the CB1 receptor rather than through it, acting instead on serotonin 5-HT1A, TRPV1, GPR55 and adenosine signalling among others — a genuinely polypharmacological profile.

Its strongest evidence is in rare epilepsies, which is what it is approved for; that is a narrow and severe indication rather than the general wellness use it is sold for. The doses that produced that evidence are also far larger than the doses most people take.

Epidiolex vs. the CBD You Can Buy

Both are cannabidiol. Almost nothing else about them is the same, and the difference is mostly one of scale.

The dose gap is one to two orders of magnitude. The epilepsy trials used purified CBD oral solution at 10–20 mg/kg/day ^{1,2,3Reference 1Devinsky O et al. · 2017RCTTrial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome — randomised controlled trialView study →Reference 2Devinsky O et al. · 2018RCTEffect of Cannabidiol on Drop Seizures in the Lennox-Gastaut Syndrome — randomised controlled trialView study →Reference 3Thiele EA et al. · 2018RCTCannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4) — randomised controlled trialView study →} — for a 70 kg adult, 700 to 1,400 mg a day. The anxiety result used a single 600 mg dose ^{4Reference 4Bergamaschi MM et al. · 2011RCTCannabidiol reduces the anxiety induced by simulated public speaking in treatment-naive social phobia patients — randomised controlled trialView study →}. The cannabis-use-disorder trial used 400–800 mg/day ^{5Reference 5Freeman TP et al. · 2020RCTCannabidiol for the treatment of cannabis use disorder: a phase 2a, double-blind, placebo-controlled, randomised trialView study →}. Retail CBD products are commonly sold at 10–50 mg per serving. A 25 mg gummy is not a small dose of the studied regimen; it is between 2% and 6% of it.

Absorption makes the gap wider, not narrower. Oral CBD bioavailability is about 6% and highly variable, and a high-fat meal can multiply exposure several-fold ^{7Reference 7Millar SA et al. · 2018Systematic reviewA Systematic Review on the Pharmacokinetics of Cannabidiol in Humans — review of human PKView study →}. So the milligram figure on a label is a poor predictor of what reaches the blood, and two people taking the same product can be at very different exposures.

The high doses are not free. In healthy adults given 1,500 mg/day, ALT rose above normal in a substantial fraction and met drug-induced-liver-injury thresholds in some, resolving on stopping ^{9Reference 9Watkins PB et al. · 2021Clinical trialCannabidiol and abnormal liver chemistries in healthy adults: results of a phase I clinical trialView study →}. CBD inhibits several cytochrome P450 enzymes and roughly tripled exposure to clobazam’s active metabolite in a controlled study ^{8Reference 8Morrison G et al. · 2019A Phase 1 Open-Label Trial of Drug-Drug Interactions Between Clobazam, Stiripentol or Valproate and Cannabidiol in Healthy Subjects — human pharmacokinetic studyView study →}. These are properties of the dose that works, which is exactly why the effective form is a prescription medicine with monitoring rather than a shelf product.

The honest position is not that retail CBD is fake. It is that the evidence belongs to a purified, high-dose, monitored preparation, and the retail product shares its molecule and very little else.

Working Around the Cannabinoid Receptor

CBD is a cannabinoid that barely uses the cannabinoid receptors, which is both why it is not intoxicating and why its pharmacology is so hard to summarise.

At CB1 — the receptor THC activates to produce the high — CBD has very low affinity and no meaningful agonism. What it does instead is distributed across a set of unrelated targets: 5-HT1A serotonin receptor agonism, TRPV1/TRPV2 vanilloid channel activation, inhibition of FAAH (which raises the body’s own anandamide rather than supplying an outside agonist), GPR55 antagonism, and PPARγ activation.

Two consequences follow.

It explains the breadth and the weakness together. A molecule with a dozen modest interactions will produce a signal in a dozen different assays, and a strong result in very few. That is precisely the shape of CBD’s evidence: overwhelming in one narrow indication, thin and preliminary everywhere else.

It explains why the anticonvulsant mechanism is still unsettled. CBD’s effect in Dravet and Lennox-Gastaut syndromes is among the best-replicated findings for any plant compound ^{1,2,3Reference 1Devinsky O et al. · 2017RCTTrial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome — randomised controlled trialView study →Reference 2Devinsky O et al. · 2018RCTEffect of Cannabidiol on Drop Seizures in the Lennox-Gastaut Syndrome — randomised controlled trialView study →Reference 3Thiele EA et al. · 2018RCTCannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4) — randomised controlled trialView study →}, and nobody can say with confidence which of its targets produces it.

What It's Used For

Cannabidiol is the non-intoxicating counterpart to THC and the dominant cannabinoid in most hemp. Its pharmacology is unusual: it barely touches the CB1 receptor that drives the cannabis “high,” and instead works through a scatter of other targets — serotonin 5-HT1A receptors, TRPV vanilloid channels, the enzyme FAAH (which raises the body’s own endocannabinoids), GPR55 and PPARγ. That broad, indirect profile is why CBD turns up in so many contexts, and why its evidence is strongest where it has been tested as a purified pharmaceutical rather than a wellness oil. One purified formulation (Epidiolex) is an approved anti-seizure medicine ^{1,2,3Reference 1Devinsky O et al. · 2017RCTTrial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome — randomised controlled trialView study →Reference 2Devinsky O et al. · 2018RCTEffect of Cannabidiol on Drop Seizures in the Lennox-Gastaut Syndrome — randomised controlled trialView study →Reference 3Thiele EA et al. · 2018RCTCannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4) — randomised controlled trialView study →}.

Evidence by IndicationStrength of support · Beta

01Drug-resistant epilepsyProven88%Pivotal placebo-controlled RCTs of purified CBD; FDA/EMA-approved (Epidiolex) ^{1,2,3Reference 1Devinsky O et al. · 2017RCTTrial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome — randomised controlled trialView study →Reference 2Devinsky O et al. · 2018RCTEffect of Cannabidiol on Drop Seizures in the Lennox-Gastaut Syndrome — randomised controlled trialView study →Reference 3Thiele EA et al. · 2018RCTCannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4) — randomised controlled trialView study →}

02AnxietyPromising50%Small RCTs at 300–600 mg in acute-stress/social-anxiety models ^{4Reference 4Bergamaschi MM et al. · 2011RCTCannabidiol reduces the anxiety induced by simulated public speaking in treatment-naive social phobia patients — randomised controlled trialView study →}

03Cannabis use disorderPromising45%One well-conducted phase 2a RCT, positive ^{5Reference 5Freeman TP et al. · 2020RCTCannabidiol for the treatment of cannabis use disorder: a phase 2a, double-blind, placebo-controlled, randomised trialView study →}

04Inflammatory bowel diseaseShaky44%RCT: symptom & quality-of-life gains, no mucosal healing ^{6Reference 6Naftali T et al. · 2021RCTOral CBD-rich Cannabis Induces Clinical but Not Endoscopic Response in Patients with Crohn’s Disease — randomised controlled trialView study →}

Mechanisms

Clinical trials

CBD has an unusually large trial footprint for a plant compound, concentrated in epilepsy (where a purified product is licensed) with a growing spread into anxiety, substance use, inflammation and pain.

Last checked: July 2026.

Pharmacokinetics

CBD’s practical weak point is absorption. A systematic review of human CBD pharmacokinetics found oral bioavailability low and highly variable, on the order of 6%, because of poor water solubility and extensive first-pass metabolism — and found that a high-fat meal can raise exposure several-fold, the so-called food effect ^{7Reference 7Millar SA et al. · 2018Systematic reviewA Systematic Review on the Pharmacokinetics of Cannabidiol in Humans — review of human PKView study →}. Reported half-life varies widely with route and regimen.

The practical upshot is that dose alone is a poor guide to what reaches the blood: formulation and whether the dose is taken with food matter as much as milligrams ^{7Reference 7Millar SA et al. · 2018Systematic reviewA Systematic Review on the Pharmacokinetics of Cannabidiol in Humans — review of human PKView study →}. It also means the trial doses cannot be scaled down proportionally — a tenth of the dose in a different formulation on an empty stomach is not a tenth of the exposure.

CBD is metabolised largely by CYP3A4 and CYP2C19, and it inhibits several P450 enzymes itself, which is where its interaction burden comes from ^{8Reference 8Morrison G et al. · 2019A Phase 1 Open-Label Trial of Drug-Drug Interactions Between Clobazam, Stiripentol or Valproate and Cannabidiol in Healthy Subjects — human pharmacokinetic studyView study →}.

Isolate vs. Plant Studies

Unusually for a plant compound, the good evidence is the isolate — and that is a strength rather than a caveat.

Every result in the table above used purified CBD except one. The epilepsy trials used a pharmaceutical oral solution of purified cannabidiol ^{1,2,3Reference 1Devinsky O et al. · 2017RCTTrial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome — randomised controlled trialView study →Reference 2Devinsky O et al. · 2018RCTEffect of Cannabidiol on Drop Seizures in the Lennox-Gastaut Syndrome — randomised controlled trialView study →Reference 3Thiele EA et al. · 2018RCTCannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4) — randomised controlled trialView study →}; the anxiety and cannabis-use-disorder trials used purified CBD at defined milligram doses ^{4,5Reference 4Bergamaschi MM et al. · 2011RCTCannabidiol reduces the anxiety induced by simulated public speaking in treatment-naive social phobia patients — randomised controlled trialView study →Reference 5Freeman TP et al. · 2020RCTCannabidiol for the treatment of cannabis use disorder: a phase 2a, double-blind, placebo-controlled, randomised trialView study →}. There is no ambiguity about what was given.

The exception is instructive. The Crohn’s disease trial used oral CBD-rich cannabis, a whole-plant preparation, and produced clinical and quality-of-life improvement without endoscopic healing ^{6Reference 6Naftali T et al. · 2021RCTOral CBD-rich Cannabis Induces Clinical but Not Endoscopic Response in Patients with Crohn’s Disease — randomised controlled trialView study →} — a result that cannot be attributed to CBD rather than to the other cannabinoids and terpenes in the extract.

What does not exist is evidence for the middle ground that most of the market occupies: “full-spectrum” hemp extracts at retail doses. Those are neither the purified high-dose pharmaceutical nor the studied whole-plant preparation, and no trial has tested them.

Its Role in Plants

CBD is not made as CBD. Cannabis synthesises cannabidiolic acid (CBDA), and heat decarboxylates it — the same chemistry described in detail on tetrahydrocannabinol.

The enzyme is CBDA synthase, purified from Cannabis sativa in 1996 and shown to catalyse the oxidocyclisation of cannabigerolic acid to cannabidiolic acid ^{13Reference 13Taura F et al. · 1996Purification and characterization of cannabidiolic-acid synthase from Cannabis sativa L.: biochemical analysis of a novel enzyme that catalyzes the oxidocyclization of cannabigerolic acid to cannabidiolic acidView study →}. It competes for that single substrate with THCA synthase and CBCA synthase, and which enzyme a plant carries and expresses is what makes it a chemotype: CBDA synthase is described as the chemotype-determining enzyme of fibre-type cannabis ^{14Reference 14Taura F et al. · 2007Cannabidiolic-acid synthase, the chemotype-determining enzyme in the fiber-type Cannabis sativaView study →}. A hemp plant is not a cannabis plant with the THC removed; it is a cannabis plant that routes the same precursor down a different branch.

Like the other cannabinoids, CBDA accumulates in the glandular trichomes of the female inflorescence, outside the living cells that make it.

Prevalence in Nature

CBD is, for practical purposes, a Cannabis sativa compound. It sits among more than a hundred catalogued phytocannabinoids from the plant ^{15Reference 15Hanuš LO et al. · 2016ReviewPhytocannabinoids: a unified critical inventory — reviewView study →}, and no other genus is an established source.

Its distribution within cannabis is the mirror image of THC’s, and it has been moving. In drug-type cannabis CBD is a trace component and shrinking: across 38,681 DEA-seized samples, CBD fell from about 0.28% of dried flower in 2001 to under 0.15% by 2014, taking the THC:CBD ratio from roughly 14:1 to roughly 80:1 as breeders selected for potency ^{10Reference 10ElSohly MA et al. · 2016ObservationalChanges in cannabis potency over the last two decades (1995–2014): analysis of current data in the United States — observationalView study →}. In the 2013–2022 follow-up, every cannabinoid other than THC sat below 0.5% in high-THC samples, apart from CBG and CBN ^{10Reference 10ElSohly MA et al. · 2016ObservationalChanges in cannabis potency over the last two decades (1995–2014): analysis of current data in the United States — observationalView study →}.

CBD-chemotype hemp is the opposite plant. Industrial cultivars bred for CBD keep THC below the 0.3% statutory ceiling while accumulating CBDA as their dominant cannabinoid; in one three-cultivar Chinese survey, CBDA content rose across the series as THC fell ^{16Reference 16Jiang X et al. · 2026Cannabinoids and terpenoids in hemp inflorescence and leaves of three varieties in Yunnan, ChinaView study →}. Published percentage figures for CBD in these cultivars are surprisingly scarce, and no consistent range exists that could be compared across provenances.

Major Natural Sources

Source	Part	Level	Note
CBD-chemotype hemp (Cannabis sativa)	dried female inflorescence	dominant cannabinoid; no consistent published range	present as CBDA before heating 14,16Reference 14Taura F et al. · 2007Cannabidiolic-acid synthase, the chemotype-determining enzyme in the fiber-type Cannabis sativaView study →Reference 16Jiang X et al. · 2026Cannabinoids and terpenoids in hemp inflorescence and leaves of three varieties in Yunnan, ChinaView study →
Drug-type cannabis (Cannabis sativa)	dried female inflorescence	~0.28% (2001) → <0.15% (2014)	falling as THC was selected up 10Reference 10ElSohly MA et al. · 2016ObservationalChanges in cannabis potency over the last two decades (1995–2014): analysis of current data in the United States — observationalView study →
Cannabis	leaf, stem	low, trichome-density dependent	no systematic survey
Cannabis	seed, root	negligible	no glandular trichomes

Discovery & Synthesis

CBD was isolated before THC and before anyone knew what it was for. Roger Adams and colleagues at the University of Illinois obtained it from an extract of Minnesota wild hemp and published its structure in 1940 ^{11Reference 11Adams R et al. · 1940Structure of cannabidiol, a product isolated from the marihuana extract of Minnesota wild hempView study →} — twenty-four years before THC was characterised, and at a time when the psychoactive principle of cannabis was still unidentified. The full structure and stereochemistry were settled in 1963 by Mechoulam and Shvo, the first paper in the series that would produce THC the following year ^{12Reference 12Mechoulam R · 1963Hashish—I: the structure of cannabidiolView study →}.

Manufacture today is extraction. Pharmaceutical CBD is purified from purpose-grown high-CBD cannabis rather than synthesised, and the same is true of the supplement supply. Total synthesis of CBD is straightforward chemistry and is done, but it does not compete on cost with a plant that makes the molecule at percent-level concentrations.

Patents: not yet researched (future patent-loop pass).

Toxicity & Safety

CBD is well tolerated, with the commonest effects being drowsiness, diarrhea and appetite change ^{1,2Reference 1Devinsky O et al. · 2017RCTTrial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome — randomised controlled trialView study →Reference 2Devinsky O et al. · 2018RCTEffect of Cannabidiol on Drop Seizures in the Lennox-Gastaut Syndrome — randomised controlled trialView study →}. The interactions matter more than the side effects. CBD inhibits several cytochrome-P450 enzymes (CYP3A4, CYP2C9, CYP2C19, CYP2D6, CYP1A2), so it can raise blood levels of co-administered drugs — a controlled study showed CBD roughly tripled exposure to the active metabolite of the epilepsy drug clobazam via CYP2C19 inhibition ^{8Reference 8Morrison G et al. · 2019A Phase 1 Open-Label Trial of Drug-Drug Interactions Between Clobazam, Stiripentol or Valproate and Cannabidiol in Healthy Subjects — human pharmacokinetic studyView study →}. And high-dose CBD can raise liver enzymes: in healthy adults given 1,500 mg/day, ALT rose above normal in nearly half and met drug-induced-liver-injury thresholds in about a third, resolving on stopping ^{9Reference 9Watkins PB et al. · 2021Clinical trialCannabidiol and abnormal liver chemistries in healthy adults: results of a phase I clinical trialView study →}. Anyone on regular medication should check for interactions before use, and the unregulated supplement market means product content varies.

Dosage

Studied doses span a wide range depending on the goal — and they sit well above typical wellness products. These are doses studied in research, not a recommendation, and high-dose use is a medical decision given the interactions above.

References

1. Devinsky O, Cross JH, Laux L, et al. (2017). Trial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome — randomised controlled trial. New England Journal of Medicine. https://pubmed.ncbi.nlm.nih.gov/28538134/
2. Devinsky O, Patel AD, Cross JH, et al. (2018). Effect of Cannabidiol on Drop Seizures in the Lennox-Gastaut Syndrome — randomised controlled trial. New England Journal of Medicine. https://pubmed.ncbi.nlm.nih.gov/29768152/
3. Thiele EA, Marsh ED, French JA, et al. (2018). Cannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4) — randomised controlled trial. The Lancet. https://pubmed.ncbi.nlm.nih.gov/29395273/
4. Bergamaschi MM, Queiroz RH, Chagas MH, et al. (2011). Cannabidiol reduces the anxiety induced by simulated public speaking in treatment-naive social phobia patients — randomised controlled trial. Neuropsychopharmacology. https://pubmed.ncbi.nlm.nih.gov/21307846/
5. Freeman TP, Hindocha C, Baio G, et al. (2020). Cannabidiol for the treatment of cannabis use disorder: a phase 2a, double-blind, placebo-controlled, randomised trial. The Lancet Psychiatry. https://pubmed.ncbi.nlm.nih.gov/32735782/
6. Naftali T, Bar-Lev Schleider L, Almog S, et al. (2021). Oral CBD-rich Cannabis Induces Clinical but Not Endoscopic Response in Patients with Crohn’s Disease — randomised controlled trial. Journal of Crohn’s & Colitis. https://pubmed.ncbi.nlm.nih.gov/33858011/
7. Millar SA, Stone NL, Yates AS, O’Sullivan SE (2018). A Systematic Review on the Pharmacokinetics of Cannabidiol in Humans — review of human PK. Frontiers in Pharmacology. https://pubmed.ncbi.nlm.nih.gov/30534073/
8. Morrison G, Crockett J, Blakey G, Sommerville K (2019). A Phase 1 Open-Label Trial of Drug-Drug Interactions Between Clobazam, Stiripentol or Valproate and Cannabidiol in Healthy Subjects — human pharmacokinetic study. Clinical Pharmacology in Drug Development. https://pubmed.ncbi.nlm.nih.gov/30791225/
9. Watkins PB, Church RJ, Li J, Knappertz V (2021). Cannabidiol and abnormal liver chemistries in healthy adults: results of a phase I clinical trial. Clinical Pharmacology & Therapeutics. https://pubmed.ncbi.nlm.nih.gov/33022751/
10. ElSohly MA, Mehmedic Z, Foster S, et al. (2016). Changes in cannabis potency over the last two decades (1995–2014): analysis of current data in the United States — observational. Biological Psychiatry. https://pubmed.ncbi.nlm.nih.gov/26903403/
11. Adams R, Hunt M, Clark JH (1940). Structure of cannabidiol, a product isolated from the marihuana extract of Minnesota wild hemp. I. Journal of the American Chemical Society. https://doi.org/10.1021/ja01858a058
12. Mechoulam R, Shvo Y (1963). Hashish—I: the structure of cannabidiol. Tetrahedron. https://doi.org/10.1016/0040-4020(63)85022-x
13. Taura F, Morimoto S, Shoyama Y (1996). Purification and characterization of cannabidiolic-acid synthase from Cannabis sativa L.: biochemical analysis of a novel enzyme that catalyzes the oxidocyclization of cannabigerolic acid to cannabidiolic acid. Journal of Biological Chemistry. https://pubmed.ncbi.nlm.nih.gov/8663284/
14. Taura F, Sirikantaramas S, Shoyama Y, et al. (2007). Cannabidiolic-acid synthase, the chemotype-determining enzyme in the fiber-type Cannabis sativa. FEBS Letters. https://pubmed.ncbi.nlm.nih.gov/17544411/
15. Hanuš LO, Meyer SM, Muñoz E, et al. (2016). Phytocannabinoids: a unified critical inventory — review. Natural Product Reports. https://pubmed.ncbi.nlm.nih.gov/27722705/
16. Jiang X, Cai Y, Teng Y, et al. (2026). Cannabinoids and terpenoids in hemp inflorescence and leaves of three varieties in Yunnan, China. Rapid Communications in Mass Spectrometry. https://pubmed.ncbi.nlm.nih.gov/42185735/

View all 16 references▾