CBD (Cannabidiol): The Dose Gap Between Medicine and the Market

CBD went from a chemist’s curiosity to a supermarket ingredient in about a decade, and the science did not keep pace with the shelf space. The good news is real: a purified CBD medicine cuts seizures in three rare, severe epilepsy syndromes, with the trial record to prove it. Prescription CBD doses are commonly measured in hundreds of milligrams and, for larger patients, grams; the gummy is measured in tens of milligrams. That dose gap runs through much of the evidence, although it is not the whole story. Millions of people take CBD for anxiety, sleep and pain, and many say it helps; consumer-sized doses remain poorly studied for many popular uses, and the limited trials conducted at or near those doses mostly did not beat placebo. The liver signal has now been confirmed in healthy adults, drug interactions are well documented, and products on the shelf often do not contain what their labels claim. This article covers what CBD is, how it works, what users report, what the trials found, and what changes in November and December 2026.

What Is CBD and Where Does It Come From?

Cannabidiol is one of the two major cannabinoids in Cannabis sativa, alongside THC. The plant does not make either directly. It makes cannabigerolic acid (CBGA), and an enzyme converts it to cannabidiolic acid (CBDA); heat or time then removes the carboxyl group and CBDA becomes CBD. Our acidic cannabinoids profile covers the first step and our decarboxylation explainer the second. Hemp grown for fiber and seed contains little of it; the “CBD hemp” planted since 2018 is bred for a CBD-dominant profile with very low THC.

CBD and THC share a formula and differ in one ring: THC’s is closed, CBD’s is open. That small difference helps explain why CBD has little direct agonist activity at CB1 and does not produce a THC-like high. Its shorter-chained sibling, CBDV, and the other minor cannabinoids, CBG, CBN and CBC, each have their own profile on this site.

Does CBD Get You High?

No, and the reason is worth getting right. CBD has weak affinity for the two cannabinoid receptors. In cell assays, it acts as a negative allosteric modulator at CB1: it binds at a site separate from THC and changes receptor signaling.^[1] Whether that translates into CBD reliably blunting THC in people is less clear. Human results depend on dose, route and formulation, and some oral studies have found that high-dose CBD increases THC exposure and effects. “Non-intoxicating” is the accurate word. “Non-psychoactive,” which appears on a lot of packaging, is looser: a compound that changes anxiety or sleepiness is acting on the mind, even without a high. For how the receptors fit together, see our endocannabinoid system pillar.

How Does CBD Work?

Mostly not through the cannabinoid receptors, which is why it took so long to pin down. In cells, CBD activates and then desensitizes TRPV1, the capsaicin channel, and weakly slows the breakdown and reuptake of anandamide, the body’s own cannabinoid.^[2] It is a modest-affinity agonist at the serotonin 5-HT1A receptor, the target most often invoked for its anti-anxiety effects.^[3] Add the allosteric action at CB1^[1] and you have a drug with several moderate actions rather than one strong one. Why that kind of pharmacology resists simple stories is a theme of our interview with pharmacologist Jahan Marcu. Every one of those findings comes from cell or animal work; which of them matters at the doses people actually take is not settled. What is settled is that the anti-seizure effect is real in people and appears not to depend on CB1.

What Has CBD Actually Been Approved For?

Three rare, severe epilepsies, on the strength of four placebo-controlled trials run by GW Pharmaceuticals (now part of Jazz). All used a purified oral CBD solution added to existing anti-seizure drugs for 14 to 16 weeks.

Those are large effects for drug-resistant epilepsy, and they came at a price: somnolence, diarrhea, reduced appetite, and liver enzyme elevations that were more common at higher doses and in children also taking valproate. In the tuberous sclerosis trial, 19 percent of patients on CBD had transaminase levels more than three times the upper limit of normal, none on placebo, and four in five of them were on valproate.^[7] The FDA approved Epidiolex for Dravet and Lennox-Gastaut in 2018 and added tuberous sclerosis in 2020. Note the doses: 10 to 50 mg per kilogram per day, which is 700 to 3,500 mg a day for a 70 kg adult. A typical gummy contains 10 to 25 mg.

Our 2017 news item on CBD and tuberous sclerosis covered an uncontrolled, company-sponsored series from Mexico; the trial above is the evidence that replaced it.

What About Anxiety, Sleep and Pain?

This is where the shelf and the literature part ways, and where it matters most to hold three things side by side: what users report, what the trials found, and at what dose.

Anxiety. It is one of the leading reasons people use CBD. In a self-selected survey of 2,409 predominantly American users, pain conditions and anxiety were among the most common reasons for use, and 36 percent of those treating a medical condition said CBD worked “very well by itself.”^[8] In a UK survey of 387 users, 43 percent took it for anxiety and 87 percent of those said they felt less anxious.^[9] Both were self-selected surveys of people already using CBD. They tell us what users perceive, not whether CBD caused the improvement. The trials are thinner. The best-known result is a single 300 to 600 mg dose reducing anxiety during a simulated public-speaking test. A 2024 systematic review found 15 placebo-controlled trials; nine found no benefit from a single dose, some higher doses appeared to worsen anxiety, and the authors could not pool the results because the designs were so different.^[10] An earlier meta-analysis pooled just two small social-anxiety studies and found a large but imprecise effect, then concluded cannabis preparations “should be considered experimental.”^[11] The largest synthesis to date, 54 randomized trials of cannabinoids for mental and substance-use disorders published in 2026, found no significant effect on anxiety or PTSD and rated the evidence low quality overall.^[12] The dose line: positive results generally came from 300 to 600 mg taken once, 12 to 60 times the amount in a typical 10 to 25 mg gummy. Daily dosing at consumer strengths has barely been tested.

Sleep. In the UK survey, sleep problems were reported almost as often as anxiety: 42 percent of the sample took CBD for sleep problems, and about half of those said it helped them fall asleep faster.^[9] Most CBD sleep claims lean on sedation seen at epilepsy doses or on products that also contain THC or CBN. Our cannabis and sleep article goes through the trials; the short version is that pure CBD has not been shown to improve sleep in a well-controlled study at the doses people buy. The dose line: somnolence is documented at prescription doses, but feeling sleepy is not the same as treating insomnia. Consumer-sized doses remain poorly tested.

Pain. Pain is the condition people most often treat with CBD on its own, and the survey respondents who used it for chronic pain and arthritis were among the most satisfied.^[8] The trial record points the other way. A 2024 review in the Journal of Pain found 16 randomized trials of pharmaceutical-grade CBD across 12 pain conditions, at doses from 6 to 1,600 mg; fifteen found no benefit over placebo. Its title was “ineffective, expensive, and with potential harms.”^[13] Since then a 200-patient Danish trial gave 50 mg of CBD or placebo daily for 24 weeks to people with fibromyalgia. Pain fell more on placebo, by 0.7 points on a 10-point scale, and the difference was statistically significant in placebo’s favor.^[14] The federal living review of cannabis for chronic pain counts only six placebo-controlled trials of CBD alone, at a median dose of about 50 mg a day, too few and too small to support a claim either way.^[15] The dose line here is the uncomfortable one: the fibromyalgia trial used a consumer dose, but the null results in the 2024 review ran all the way from 6 to 1,600 mg. For pain, the dose gap does not rescue CBD.

Psychosis. One signal runs the other way, and it fits the dose story. In 88 patients with schizophrenia already on antipsychotics, six weeks of 1,000 mg CBD a day reduced positive symptoms more than placebo, and clinicians rated 79 percent of the CBD group as improved, compared with 55 percent receiving placebo.^[16] It is a single trial at a gram a day, the 2026 meta-analysis found no significant pooled effect in psychotic disorders, and it needs replication; but it is a real result at a real dose, and it is the kind of study the consumer market has never funded.

None of this says CBD does nothing in people. It says that users commonly report benefits at consumer-sized doses that remain poorly studied; that larger, prescription-style doses have produced clear effects for some conditions and null results for others; and that where rigorous trials have tested consumer-sized doses, as in fibromyalgia, placebo has so far won.

Is CBD Safe?

CBD is generally well tolerated in short-term studies, but it is not inert. Three caveats are now well documented.

The liver. Transaminase elevations showed up in every Epidiolex trial. In 2025 a randomized trial published in JAMA Internal Medicine tested whether that was a high-dose artifact: 201 healthy adults took purified CBD at 5 mg per kilogram per day, about 350 mg for a 70 kg person, or placebo for four weeks. Eight people receiving CBD, and none receiving placebo, developed liver enzyme levels more than three times the upper limit of normal. Seven met the study’s withdrawal criteria for potential drug-induced liver injury, although that designation did not mean they had developed clinically apparent liver injury, and the levels came down when the CBD stopped.^[17] That dose is commercially available and near the upper end of reported consumer use, far above a typical gummy.

Drug interactions. CBD is metabolized by, and inhibits, the same liver enzymes that handle many prescription drugs. In 81 epilepsy patients, adding CBD pushed the active metabolite of clobazam above the therapeutic range and raised levels of several other anti-seizure drugs, and liver enzymes rose in people also taking valproate.^[18] Anyone on medication with a narrow margin should treat CBD as a drug, not a supplement.

Regulatory status. In January 2023 the FDA concluded that the existing regulatory frameworks for foods and dietary supplements are “not appropriate” for CBD, citing liver effects, drug interactions, possible effects on the male reproductive system, and the absence of safety data in children, the elderly and pregnant women, and said a new pathway would require Congress.^[19] The WHO’s 2018 review found no abuse or dependence potential and recommended against international control.^[20] Both are true; they answer different questions.

How Much CBD Gets Into You?

Less than you think, and it depends on breakfast. Oral CBD is poorly absorbed and heavily metabolized in the liver; the systematic review of human pharmacokinetics could not find a reliable absolute figure for oral bioavailability, only a 31 percent value for smoking.^[21] What is firmly established is the food effect. In the Phase 1 trials behind Epidiolex, a high-fat meal raised peak CBD blood levels nearly fivefold and total exposure about fourfold.^[22] That result came from high doses of a purified oral solution, so the exact multiplier should not be assumed for every gummy or oil. Loren DeVito’s CBD dosing and the food effect covers the companion study in patients. The practical point remains: CBD exposure depends strongly on food and formulation, and the same product may deliver substantially more CBD after a high-fat meal than in the fasted state.

What Is Actually in the Bottle?

Often not what the label says. When researchers bought 84 CBD extracts online and tested them, only 31 percent were accurately labeled; 43 percent contained more CBD than the label stated, 26 percent contained less, and 21 percent contained detectable THC.^[23] The 2024 pain review found the same picture years later: products ranging from no CBD to well above the labeled amount, with contaminants.^[13] A recent, batch-specific certificate of analysis from an accredited laboratory is the best available check, although it is only as reliable as the sample and laboratory behind it; our cannabis testing guide explains how to read one, and our page on lab testing requirements explains why hemp-derived CBD escapes most of the testing that state-legal cannabis must pass. For the consumer angle, our sister publication CBD Health & Wellness covers products and buying.

Is CBD Legal?

In the United States, CBD derived from cannabis containing no more than 0.3 percent delta-9 THC has generally fallen within the federal definition of hemp since 2018 and is therefore excluded from the Controlled Substances Act. That does not automatically make every CBD product lawful: FDA requirements and state laws still apply. That definition is now changing in two stages. On November 12, 2026, cannabinoids not naturally produced by the plant fall out of the hemp definition; on December 11, 2026, the standard shifts to total THC, including THCA, with a cap of 0.4 mg of THC per container for finished products.^[24] Pure CBD products are not the target, but full-spectrum extracts that carry meaningful THC are. Our hemp redefinition explainer has the detail. None of this changes the FDA position above: legal to grow and sell is not the same as approved to put in food or call a supplement.

The Bottom Line

CBD is a real drug. It has four placebo-controlled trials and an approved medicine behind it, a plausible signal in psychosis, and a safety profile that is good but not blank: at hundreds of milligrams a day it can elevate liver enzymes, and it can alter the levels of other medicines. The problem is that it is sold like a supplement. CBD unquestionably works for certain epilepsies at prescription doses. Outside epilepsy, there is a suggestive result in schizophrenia, a few positive anxiety experiments and a much larger body of uncertain or negative evidence. Consumer-sized doses remain poorly studied for many popular uses, but increasing the dose does not automatically make CBD effective: the pain trials were largely negative across doses from 6 to 1,600 mg. That is not a reason to feel foolish for taking it; many people report that it helps. If you take it, buy a product with a current certificate of analysis, use it under consistent food conditions, and tell your doctor if you take prescription medicines. Decide beforehand what improvement you are looking for, and do not assume that taking more will work better if no meaningful benefit appears.

References

1. Laprairie RB, Bagher AM, Kelly MEM, Denovan-Wright EM. Cannabidiol is a negative allosteric modulator of the cannabinoid CB1 receptor. British Journal of Pharmacology. 2015;172(20):4790–4805.
2. Bisogno T, Hanuš L, De Petrocellis L, et al. Molecular targets for cannabidiol and its synthetic analogues: effect on vanilloid VR1 receptors and on the cellular uptake and enzymatic hydrolysis of anandamide. British Journal of Pharmacology. 2001;134(4):845–852.
3. Russo EB, Burnett A, Hall B, Parker KK. Agonistic properties of cannabidiol at 5-HT1a receptors. Neurochemical Research. 2005;30(8):1037–1043.
4. Devinsky O, Cross JH, Laux L, et al. Trial of cannabidiol for drug-resistant seizures in the Dravet syndrome. New England Journal of Medicine. 2017;376:2011–2020.
5. Thiele EA, Marsh ED, French JA, et al. Cannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4): a randomised, double-blind, placebo-controlled phase 3 trial. The Lancet. 2018.
6. Devinsky O, et al. Effect of cannabidiol on drop seizures in the Lennox-Gastaut syndrome. New England Journal of Medicine. 2018;378:1888–1897.
7. Thiele EA, Bebin EM, Bhathal H, et al. Add-on cannabidiol treatment for drug-resistant seizures in tuberous sclerosis complex: a placebo-controlled randomized clinical trial. JAMA Neurology. 2021.
8. Corroon J, Phillips JA. A cross-sectional study of cannabidiol users. Cannabis and Cannabinoid Research. 2018;3(1):152–161.
9. Moltke J, Hindocha C. Reasons for cannabidiol use: a cross-sectional study of CBD users, focusing on self-perceived stress, anxiety, and sleep problems. Journal of Cannabis Research. 2021;3:5.
10. Coelho CDF, Vieira RP, Araújo-Junior OS, et al. The impact of cannabidiol treatment on anxiety disorders: a systematic review of randomized controlled clinical trials. Life. 2024;14(11):1373.
11. Bahji A, Meyyappan AC, Hawken ER. Efficacy and acceptability of cannabinoids for anxiety disorders in adults: a systematic review and meta-analysis. Journal of Psychiatric Research. 2020.
12. Wilson J, Dobson O, Langcake A, et al. The efficacy and safety of cannabinoids for the treatment of mental disorders and substance use disorders: a systematic review and meta-analysis. The Lancet Psychiatry. 2026;13(4):304–315.
13. Moore A, Straube S, Fisher E, Eccleston C. Cannabidiol (CBD) products for pain: ineffective, expensive, and with potential harms. Journal of Pain. 2024;25(4):833–842.
14. Rasmussen MU, Christensen R, Wæhrens EE, et al. Cannabidiol versus placebo in patients with fibromyalgia: a randomised, double-blind, placebo-controlled, parallel-group, single-centre trial. Annals of the Rheumatic Diseases. 2026.
15. Chou R, Ahmed AY, Dana T, et al. Living systematic review on cannabis and other plant-based treatments for chronic pain: 2025 update. Agency for Healthcare Research and Quality, Comparative Effectiveness Review No. 250; July 2025.
16. McGuire P, Robson P, Cubala WJ, et al. Cannabidiol (CBD) as an adjunctive therapy in schizophrenia: a multicenter randomized controlled trial. American Journal of Psychiatry. 2018;175(3):223–231.
17. Florian J, Salcedo P, Burkhart K, et al. Cannabidiol and liver enzyme level elevations in healthy adults: a randomized clinical trial. JAMA Internal Medicine. 2025.
18. Gaston TE, Bebin EM, Cutter GR, Liu Y, Szaflarski JP. Interactions between cannabidiol and commonly used antiepileptic drugs. Epilepsia. 2017;58(9):1586–1592.
19. U.S. Food and Drug Administration. FDA concludes that existing regulatory frameworks for foods and supplements are not appropriate for cannabidiol, will work with Congress on a new way forward. Statement, January 26, 2023.
20. World Health Organization. WHO Expert Committee on Drug Dependence review of cannabis: questions and answers. July 31, 2018 (critical review of cannabidiol, June 2018).
21. Millar SA, Stone NL, Yates AS, O’Sullivan SE. A systematic review on the pharmacokinetics of cannabidiol in humans. Frontiers in Pharmacology. 2018;9:1365.
22. Taylor L, Gidal B, Blakey G, Tayo B, Morrison G. A phase I, randomized, double-blind, placebo-controlled, single ascending dose, multiple dose, and food effect trial of the safety, tolerability and pharmacokinetics of highly purified cannabidiol in healthy subjects. CNS Drugs. 2018.
23. Bonn-Miller MO, Loflin MJE, Thomas BF, et al. Labeling accuracy of cannabidiol extracts sold online. JAMA. 2017;318(17):1708–1709.
24. Congressional Research Service. Changes to the statutory definition of hemp and implications for agricultural policy. In Focus IF13136, updated September 28, 2026.

Editor’s note: This article was originally published on July 9, 2018, by Heather Ritchie. In October 2026 the editors rewrote it to cover the completed clinical trials, the 2025 liver safety data, and the 2026 change in the federal hemp definition. This article is for information only and is not medical advice.

Reviewed by Chana Frenkel, Ph.D.

Last updated October 2026 · Originally published July 9, 2018.