Search for CBD online and you will find it described as helpful for almost everything. The research does not support most of that. It also doesn't support the opposite view that CBD is "just a placebo". The truthful picture is in between, and it varies a great deal by topic. This page sorts the research by the strength of the evidence, labels every finding as human or preclinical, and explains how to read a CBD study for yourself.
First: how to weigh a CBD study
Not all studies are equal. Roughly from strongest to weakest:
- Systematic reviews and meta-analyses of several randomised trials.
- Randomised controlled trials (RCTs) in people, ideally double-blind and placebo-controlled.
- Open-label studies and case series in people, with no placebo group.
- Surveys of people who already use CBD (prone to bias: people who feel no effect tend to stop).
- Preclinical research in cells or animals.
Also check the dose. Many positive human studies used 300 to 600 mg of purified CBD in a single dose, or even more per day. That is many times the amount in a typical serving of a consumer oil or gummy. A result at 600 mg tells you little about 20 mg.
The evidence map at a glance
| Area | Best evidence | What researchers observed | Strength |
|---|---|---|---|
| Rare childhood epilepsies | Multiple large RCTs human | Fewer seizures with high-dose purified CBD on prescription | Strong (prescription medicine only) |
| Anxiety in a test situation | Small RCTs human | Lower anxiety scores during simulated public speaking at 300–600 mg | Moderate for that setting; limited beyond it |
| Everyday anxiety and stress | Small trials, case series human | Mixed; some improvement in uncontrolled studies | Weak |
| Sleep | Small trials, case series human | Mixed; improvements often secondary to anxiety changes | Weak |
| Pain | Small RCTs, reviews human | Inconsistent; many trials combine CBD with THC | Weak and mixed |
| Inflammation, immune effects | Cells and animals preclinical | Effects on inflammatory signalling in lab models | Early |
| Skin (topicals) | Lab studies, a few small human studies | Effects on skin cell models; little controlled human data | Early |
Epilepsy: the strongest evidence, and why it doesn't transfer to supplements
The best evidence for CBD comes from several large, double-blind trials in two rare and severe childhood epilepsies, Dravet syndrome and Lennox-Gastaut syndrome, and later tuberous sclerosis complex human. In those trials, researchers observed fewer seizures with purified CBD compared with placebo. This led to a licensed prescription medicine, used under specialist supervision.
Two points matter. First, the doses were high: typically 10 to 20 mg per kilogram of body weight per day, which for an adult would be well over 700 mg daily. Second, these trials also found side effects at those doses, including sleepiness and changes in liver enzyme tests. Consumer CBD products are not a substitute for prescribed epilepsy treatment, and anyone on anti-seizure medication should never add CBD without their neurologist.
Anxiety: promising in specific settings
A frequently cited study (Bergamaschi and colleagues, 2011) gave 600 mg of CBD or placebo to people with social anxiety before a simulated public-speaking test human. Researchers observed lower anxiety and discomfort scores in the CBD group. Later studies by Brazilian groups using similar test settings found an inverted-U pattern: 300 mg seemed to work better than either 150 mg or 600 mg in some experiments.
Studies of everyday anxiety are weaker. A 2019 case series from a US psychiatric clinic (Shannon and colleagues) reported that most of 72 adults had lower anxiety scores after a month of CBD alongside usual care human, but there was no placebo group, and doses were mostly 25 to 75 mg. Several placebo-controlled trials are under way or recently published with mixed results.
The honest summary: there is reasonable evidence that high single doses can reduce anxiety in a stressful test situation. There is not yet strong evidence that typical consumer doses help everyday anxiety.
Sleep: mostly indirect
Many people buy CBD for sleep. Human studies are small, and where improvements were observed they often tracked improvements in anxiety rather than sleep directly human. Some small trials found no clear difference from placebo on objective sleep measures. Some research also suggests CBD may be mildly alerting at lower doses and sedating at higher ones. Overall, the evidence is weak and mixed.
Pain: inconsistent, and often confused with THC
Pain research is harder to read than it looks, because many of the positive trials used products combining THC and CBD, not CBD alone. Systematic reviews of CBD-only products have generally found the evidence too limited or inconsistent to draw conclusions human. In animal models, researchers have observed effects on pain and inflammation pathways preclinical, which is why research continues.
Inflammation, immunity and other areas
Much of the excitement about CBD comes from lab and animal studies preclinical, where researchers observed effects on inflammatory signalling, immune cells, oxidative stress and nerve cells. These findings are genuinely interesting to scientists, but most such early results do not carry over to people. Treat headlines based on mouse or cell studies as "worth researching", not "proven".
Why studies often disagree
- Doses vary hugely, from 10 mg to over 1,000 mg a day.
- Products vary: isolate, full spectrum, with or without THC.
- Absorption varies between people and with food.
- Studies are small, often with 20 to 60 participants.
- Funding and publication bias are real in a commercial field. That includes us, which is why we link sources.
What this means for you
If you choose to try CBD, do it with realistic expectations: as part of a routine, not as a treatment. Choose a product with an independent lab report, start with a low amount, keep a simple diary of how you feel, and talk to your pharmacist first if you take any medication. Read our safety and interactions guide before you start.
How to read a CBD headline
When you see a news story or advert about CBD research, ask four quick questions:
- Was it in people? If the study was in mice or cells, it is early research, however exciting the headline.
- Was there a placebo group? Without one, it is hard to separate CBD's effect from expectation, natural improvement over time, or other treatments.
- How many people took part? A study with 20 people can point in a direction but cannot settle a question.
- What dose and product? A result from 600 mg of purified CBD cannot be applied to a 10 mg gummy, and a THC-CBD medicine is not the same as a CBD-only product.
Applying these four questions will filter out most misleading claims in minutes.
Sources
- Devinsky O et al. Trial of cannabidiol for drug-resistant seizures in the Dravet syndrome. New England Journal of Medicine, 2017.
- Thiele EA et al. Cannabidiol in patients with seizures associated with Lennox-Gastaut syndrome (GWPCARE4). The Lancet, 2018.
- Bergamaschi MM et al. Cannabidiol reduces the anxiety induced by simulated public speaking in treatment-naïve social phobia patients. Neuropsychopharmacology, 2011.
- Linares IM et al. Cannabidiol presents an inverted U-shaped dose-response curve in a simulated public speaking test. Brazilian Journal of Psychiatry, 2019.
- Shannon S et al. Cannabidiol in anxiety and sleep: a large case series. The Permanente Journal, 2019.
- Larsen C, Shahinas J. Dosage, efficacy and safety of cannabidiol administration in adults: a systematic review of human trials. Journal of Clinical Medicine Research, 2020.
Frequently asked questions
Is there proof that CBD works?
It depends on the topic. The strongest evidence is for high-dose prescription CBD in specific rare epilepsies. For anxiety, sleep and pain, human evidence is limited or mixed, and consumer CBD products cannot claim to treat any condition.
What does preclinical mean?
Preclinical research is done in cells or animals. It helps scientists decide what to test in people, but most preclinical findings do not carry over to humans.
Why do CBD studies use such high doses?
Many trials used 300 to 600 mg or more of purified CBD to see whether any effect exists at all. Typical consumer servings are much lower, so results do not transfer directly.
Does CBD help with sleep?
The evidence is weak and mixed. Where improvements were seen, they often followed changes in anxiety rather than sleep itself.
Is CBD better than a placebo?
In some well-designed trials, yes, particularly high-dose trials in epilepsy and test-situation anxiety. In others, no clear difference was found.
Can I trust CBD research funded by companies?
Look at study design, size and whether it was placebo-controlled and peer-reviewed. Industry funding doesn't make a study wrong, but independent replication matters.
Where can I read the original studies?
Each guide on cannabidiol.ie lists its main sources. Most can be found by searching the title on PubMed or Google Scholar.
Is CBD research being done in Ireland?
Yes. Irish universities have contributed to cannabinoid and pain research, mostly in the lab so far. We plan a separate guide to Irish CBD research.