Does CBD Improve Sleep Quality? What Scientific Studies Say

Sleep is one of the most common reasons people turn to cannabidiol, yet it is also the area with the weakest evidence. We reviewed the studies referenced by Polish guides and compared them with what these studies actually measured. The discrepancies were systematic. The study cited as the strongest evidence had no placebo group and its abstract did not specify any dose. The experiment cited as proof of sleep architecture safety used only one dose, not a range. A third study, cited as evidence of efficacy, showed the opposite of what the popular advice to increase the dose until effective suggests. Separately, we checked what the European Food Safety Authority wrote about cannabidiol safety in 2026, as its position changes the conversation about milligrams more than any clinical study. Below we write what is known, what is unknown, and why cognitive-behavioral therapy remains the first-line treatment for chronic insomnia rather than any preparation.

KEY INFORMATION
• Shannon 2019: sleep improved in 66.7 percent of 72 patients in the first month, but results fluctuated in subsequent months and the study had no placebo group.
• No effective CBD dose for sleep has been established; EFSA derived a provisional safe dose around 2 mg per day for a 70 kg person in 2026.
• CBD safety cannot be established for people under 25 years old, pregnant or breastfeeding women, and those taking medications.
• Cognitive-behavioral therapy for insomnia is the first-line treatment with a strong recommendation.

Do clinical studies confirm that CBD improves sleep?

Not to a degree that would allow calling it evidence. The most frequently cited work is a retrospective chart review from a psychiatric clinic, and the only polysomnographic placebo-controlled study showed no effect. These two results circulate together, although they measure different things and have very different evidential weight.

In the chart review by Shannon et al., monthly assessments of anxiety and sleep quality were recorded for 103 adult patients. The final analysis included 72 people: 47 presenting mainly for anxiety and 25 for poor sleep. Anxiety scores decreased in the first month for 57 people (79.2 percent) and remained lowered. Sleep scores improved in the first month for 48 people (66.7 percent) but fluctuated over time. Cannabidiol was well tolerated by all except three patients (Shannon et al., The Permanente Journal, 2019).

The authors conclude cautiously that cannabidiol may have value in anxiety-related disorders and that controlled clinical trials are needed. This is a more cautious statement than headlines citing this work suggest.

A 2021 knowledge summary states similarly. Much preclinical and clinical data come from small studies with assessment limitations, and clinical trials lack allocation concealment. Overall, results indicate a possible therapeutic role for cannabinoids in some sleep disorders, but authors emphasize the need for rigorous, large multicenter studies (Kaul et al., Neurotherapeutics, 2021).

What does the Shannon 2019 study not say?

It does not specify dose, has no control group, and does not measure sleep instrumentally. This is worth detailing because specific milligrams have been derived in Polish texts from this work, which it does not contain. A comparison of what the study contains versus what is attributed to it looks like this:

The lack of a placebo group is especially important because sleep is susceptible to expectations. A patient who knows they started a new therapy rates their own sleep differently than a week earlier, and the study measured self-assessment. The authors do not hide this limitation; it is only the texts citing percentages from this work that do.

Another point to remember is the asymmetry of results. The anxiety effect was stable throughout the observation period, while the sleep effect fluctuated month to month. If cannabidiol helps sleep here, it is probably indirectly, by calming anxiety, not by a sedative effect. This distinction explains why secondary insomnia due to tension responds differently than primary insomnia.

Does CBD change sleep architecture?

In the only polysomnographic placebo-controlled study, it did not change it at all. Twenty-seven healthy people received cannabidiol 300 mg or placebo in a crossover double-blind design, 30 minutes before an eight-hour polysomnographic recording. The preparation caused no significant effect, and the authors state that unlike benzodiazepines and SSRIs, cannabidiol does not disrupt the sleep cycle in healthy volunteers (Linares et al., Frontiers in Pharmacology, 2018).

This result has two sides, both worth mentioning. The good side: at this dose, cannabidiol did not alter polysomnographic recordings, so it does not disrupt the natural course of the night. The bad side: since it changed nothing in healthy people, it is not a pharmacological hypnotic. The authors explicitly call for testing this in patient populations, as their study did not include them.

A literature review on cannabis and sleep clarifies the difference from THC. Cannabidiol may have therapeutic potential in insomnia and shows promise in REM behavior disorder and excessive daytime sleepiness. THC can shorten sleep onset latency but may worsen sleep quality with prolonged use (Babson et al., Current Psychiatry Reports, 2017). The authors conclude that cannabis and sleep research is in its infancy and yields mixed results.

The closest clinical evidence comes from a case series of four patients. In Parkinson’s disease patients with REM behavior disorder, cannabidiol administration led to rapid and marked reduction in event frequency without adverse effects (Chagas et al., Journal of Clinical Pharmacy and Therapeutics, 2014). However, four cases are a case series, not a controlled study. We discuss cannabis effects on sleep more broadly in a separate post: cannabis effects on sleep.

How does CBD act on sleep-related receptors?

Contrary to most product descriptions. Cannabidiol is often presented as a CB1 cannabinoid receptor blocker similar to rimonabant, a drug withdrawn due to psychiatric effects. A systematic review of mechanisms resolves this differently.

The authors collected in vitro and ex vivo studies, synthesized them in a meta-analysis, and compared with preclinical and clinical results. The conclusion is that neither cannabidiol nor tetrahydrocannabivarin act like rimonabant, and it is very unlikely they cause central adverse effects. Cannabidiol is described as a very low affinity CB1 receptor ligand that can influence receptor activity indirectly in the body (McPartland et al., British Journal of Pharmacology, 2015).

This finding has practical significance. Since cannabidiol does not act by direct CB1 binding, there is no basis to expect a sedative effect typical of psychoactive substances. The authors caution that in vitro mechanism studies do not always predict in vivo pharmacology, so conclusions about action on a target require verification in living organisms.

Anxiolytic effects have been documented in humans. In a public speaking test, sixty healthy people aged 18 to 35 were assigned to five groups. Cannabidiol 300 mg reduced subjective anxiety scores post-performance, similarly to clonazepam, though clonazepam had stronger sedative effects. This is likely the pathway through which cannabidiol affects sleep: via tension, not the sleep center.

What CBD dose affects sleep?

It has not been established, and one well-designed study shows more is not better. In the same public speaking experiment, participants received placebo, 1 mg clonazepam, or cannabidiol at 100, 300, or 900 mg. Anxiety was reduced only by 300 mg; 100 mg and 900 mg did not differ significantly from placebo. The authors summarized this as an inverted U-shaped dose-response curve (Zuardi et al., Frontiers in Pharmacology, 2017).

This finding invalidates the most popular internet advice to gradually increase the dose until effective. If the curve is an inverted U, exceeding the optimum weakens the effect instead of strengthening it, and the user interprets this as tolerance and increases the dose further. Additionally, all these values come from anxiety studies, not sleep, so applying them to insomnia is extrapolation.

The second reference point is official and goes the opposite way. The EFSA panel recalculated subchronic study data by benchmark dose method with an uncertainty factor of 400 and derived a provisional safe dose of 0.0275 mg per kilogram body weight per day, about 2 mg for a 70 kg person (EFSA, EFSA Journal, 2026). This value applies only to supplements with cannabidiol purity of at least 98 percent, without nanoparticles.

The panel adds a sentence that should appear in every dosing text: cannabidiol safety cannot be established for people under 25 years old, pregnant or breastfeeding women, and those taking medications concurrently. For this reason, we do not provide any self-administration dosing scheme in this article. All numbers above describe specific studies, not reader recommendations; the dose is set by a doctor who knows your medications and results.

Which CBD forms are best studied?

None are well studied, and bioavailability differences listed in product tables mostly do not come from measurement. A systematic review of cannabidiol pharmacokinetics in humans examined 792 publications and found 24 with parameters measured in humans. This is the entire basis we have.

The review paints a much more modest picture than marketing. The only measured absolute bioavailability is for inhalation, at 31 percent. For other administration routes, no study attempted to establish it despite availability of intravenous preparations. Half-life ranges from 1.4 to 10.9 hours after sublingual spray and reaches 2-5 days after chronic oral administration, with time to maximum concentration between zero and four hours (Millar et al., Frontiers in Pharmacology, 2018).

Two observations from this review have practical implications for evening routine. Maximum concentration rises after a meal and in fat-based preparations, so the same number of drops acts differently on an empty stomach and after dinner. Area under the curve and maximum concentration increase with dose, and after inhalation the peak occurs earlier than after oral or sublingual routes.

Choice of preparation form is thus less decisive than comparisons suggest, and more a matter of convenience and repeatability. In the store you will find both hemp oils and swallowable forms; they differ mainly in onset speed. We discuss how minor cannabinoids fit in this context in a separate text: cannabis and sleep, CBD and CBN.

Does CBD help with sleep apnea and restless legs syndrome?

No, and this is especially important for sleep apnea because delay in causal treatment is a real risk. Data circulating in this context concern dronabinol, a synthetic THC, not cannabidiol.

In a proof-of-concept study, seventeen adults with baseline apnea-hypopnea index of at least 15 per hour received dronabinol with weekly dose escalation from 2.5 mg to 5 mg to 10 mg once daily. The change from baseline to night 21 was significant, -14.1 with a standard deviation of 17.5. No worsening of sleep architecture or serious adverse events were noted, and authors recommended confirmation in a larger study (Prasad et al., Frontiers in Psychiatry, 2013).

Cannabidiol itself has not been tested for this indication. A cannabis and sleep literature review notes that short-term benefit in apnea is suggested by synthetic cannabinoids like nabilone and dronabinol via serotonin-dependent apnea modulation. This is not the same substance sold as sleep oil.

For restless legs syndrome, the evidence base is even narrower. The available publication is a case series of six patients published in Sleep Medicine (Megelin and Ghorayeb, Sleep Medicine, 2017). A case series without control group cannot determine whether the observation results from cannabidiol, THC, or general calming. In both disorders, treatment is determined by a doctor, and sleep apnea requires prior sleep study.

How does CBD compare to cognitive-behavioral therapy?

It ranks outside first-line treatment, and this is not just our opinion. The American College of Physicians guidelines from 2016, based on a systematic review of randomized trials, recommend cognitive-behavioral therapy for insomnia as initial treatment for all adult patients with chronic insomnia. The recommendation is strong with moderate quality evidence (Qaseem et al., Annals of Internal Medicine, 2016).

Pharmacotherapy appears only in the second recommendation and in weaker form. Clinicians should use shared decision-making, discussing benefits, harms, and costs of short-term drug use, and consider adding pharmacotherapy only if cognitive-behavioral therapy fails. The recommendation is weak and evidence quality low.

Melatonin, often compared to cannabidiol, has better evidence but a small effect. A meta-analysis of nineteen studies with 1683 participants showed sleep onset shortened by 7.06 minutes and total sleep time increased by 8.25 minutes, with significant overall sleep quality improvement. Authors state the absolute benefit is less than other insomnia drugs, and melatonin is justified by a mild side effect profile (Ferracioli-Oda et al., PLOS ONE, 2013).

These three form a simple hierarchy. Cognitive-behavioral therapy has a strong recommendation, melatonin has a measured though modest effect, and cannabidiol has neither. More about the therapy and other approaches is in the post CBD in treating insomnia.

What side effects and interactions does CBD have?

They are real and dose-dependent, with the most serious involving the liver and drug interactions. The first meta-analysis of randomized placebo-controlled trials included twelve trials and 803 participants. Cannabidiol was associated with higher risk of study withdrawal and serious adverse events, with separate signals for abnormal liver tests, pneumonia, diarrhea, and sedation (Chesney et al., Neuropsychopharmacology, 2020).

A caveat from the same work changes the picture. Associations with liver tests and sedation were limited to pediatric epilepsy studies where cannabidiol may interact with clobazam or sodium valproate. Excluding these studies, diarrhea remained the only persistent effect. Sedation attributed to cannabidiol thus has a weaker basis than product descriptions for sleep suggest.

Regarding interactions, the picture is clear. A review of registration data states nearly half of cannabidiol users experienced adverse effects, and given its impact on drug metabolism targets CYP3A4 and CYP2C19 and P-glycoprotein, the potential for interactions with commonly used drugs is high (Brown and Winterstein, Journal of Clinical Medicine, 2019).

The most cited high-dose study concerns epilepsy, not sleep. One hundred twenty children and young adults with Dravet syndrome received cannabidiol 20 mg/kg/day or placebo added to standard antiepileptic treatment. Median monthly seizure frequency dropped from 12.4 to 5.9 in the cannabidiol group and from 14.9 to 14.1 in placebo, with more frequent adverse events in the treated group including abnormal liver tests (Devinsky et al., New England Journal of Medicine, 2017).

Who should not use CBD for sleep?

Three groups are explicitly named by EFSA, in a stronger form than a usual warning. The panel does not say cannabidiol is harmful in these groups, only that safety cannot be established based on available data. This applies to people under 25, pregnant and breastfeeding women, and those taking medications concurrently.

Specific observations support this. Pharmacokinetic studies confirmed cannabidiol crosses the placenta and accumulates in the body. Reproductive toxicity studies in animals raised concerns about this endpoint, and prenatal exposure showed neurodevelopmental effects suggesting long-term and sex-dependent consequences.

The document also lists two areas ignored by product descriptions. Endocrine disorders were noted, including altered thyroid hormone levels and histopathological changes in adrenal glands. Immunotoxicity has not been studied despite cannabidiol’s interaction with immune pathways, so the panel recommends caution due to lack of data rather than study results.

For this sleep article, one practical consequence is that many people seeking insomnia preparations also take medications: antidepressants, anxiolytics, thyroid drugs, or blood pressure meds. This is exactly the group EFSA says safety cannot be assessed for, and the interaction review rates the potential for drug level disruption as high. Talking with your doctor is not a formality but a condition for a sensible decision.

Is CBD addictive and does tolerance develop?

In a study on abuse potential, cannabidiol behaved like placebo. Thirty-one healthy frequent marijuana users received oral cannabidiol in a within-subject, randomized, double-blind design at 0, 200, 400, or 800 mg, alone and combined with smoked marijuana.

Active marijuana reproducibly produced subjective abuse-related effects including intoxication. Cannabidiol did not differ from placebo in any measure, and authors concluded no signals of abuse potential at tested doses (Babalonis et al., Drug and Alcohol Dependence, 2017).

Tolerance is different and requires honest assessment. We found no study tracking loss of cannabidiol’s sleep effect over time, so circulating percentages of users losing effect after weeks have no publication support. Pharmacokinetics show that with chronic oral dosing, half-life reaches several days, so the substance accumulates differently than after a single dose.

Lack of pharmacological addiction potential does not exclude attachment to the ritual itself. Evening repeated activity becomes a signal for the body, a mechanism described and used by cognitive-behavioral therapy. It is worth periodically checking whether the preparation still changes anything or has become just part of the evening.

When to see a doctor instead of trying CBD?

Always when insomnia has a red flag, and always when it is chronic. Chronic insomnia is a medical diagnosis with strongly recommended treatment, so postponing diagnosis in favor of supplements wastes time for effective therapy.

• Snoring with breathing pauses noticed by a bed partner. This suggests sleep apnea, diagnosed by sleep study, not supplement trial.
• Excessive daytime sleepiness or falling asleep at the wheel. This is urgent due to road safety.
• Suicidal thoughts, loss of interest, deeply depressed mood. First contact should be a psychiatrist, not a store.
• Nightmares replaying trauma. This signals post-traumatic stress disorder and specialist treatment.
• Insomnia with weight loss, night sweats, or shortness of breath. Somatic causes must be excluded.
• Insomnia in a person taking chronic medications. EFSA cannot assess cannabidiol safety in this group, so the decision belongs to the treating doctor.

Cannabidiol can be considered as a supportive element after consultation, mostly where insomnia has a clear anxiety component. Even then, first steps are behavioral: fixed wake-up time, limiting caffeine in the afternoon, morning daylight exposure, and evenings without screen work.

Can you drive in the morning after an evening CBD dose?

Pure cannabidiol did not impair driving in a road test, but the conclusion cannot be directly applied to full-spectrum products. In a randomized public road driving study, 26 occasional cannabis users vaporized four preparations, each with 13.75 mg of each compound. Lane deviation increased significantly after THC-dominant and balanced preparations, but not after cannabidiol-dominant, which did not differ from placebo (Arkell et al., JAMA, 2020).

The authors add a caveat often lost in abstracts: the effect size for the cannabidiol-dominant preparation did not exclude clinically significant impairment, and tested doses may not reflect typical use. After four hours, none of the conditions differed significantly from placebo.

Regarding regulations, the claim that Poland has zero THC tolerance for drivers is inaccurate. Driver testing for substances acting like alcohol is based on Article 129j of the Road Traffic Act, and the 1 nanogram THC per milliliter blood sometimes cited as a responsibility threshold is the detection limit of the method in implementing regulations. For urine, the act lists 20 nanograms per milliliter, and there is no threshold for saliva (Dz.U. 2014 poz. 948).

For evening users, the practical conclusion is simple. Full-spectrum products contain up to 0.3 percent THC calculated as the sum of delta-9-THC and tetrahydrocannabinolic acid, so THC is still present. Professional drivers find it easier to choose THC-free products with certificate of analysis than to explain trace results later.

Frequently Asked Questions

Does CBD really improve sleep quality?

The evidence is preliminary. In the retrospective chart review by Shannon 2019, sleep improved in the first month for 66.7 percent of 72 patients, but results fluctuated in subsequent months and the study had no placebo group. In the only polysomnographic placebo-controlled study, no significant effect was observed.

How much CBD should be taken for sleep?

An effective dose has not been established and this article should not be treated as a self-administration guide. EFSA derived a provisional safe dose around 2 mg per day for a 70 kg person in 2026 and stated that safety cannot be established for people taking medications. The dose is determined by a doctor.

Does CBD disrupt REM sleep?

In a polysomnographic study on healthy volunteers, a single 300 mg dose caused no significant effect on the sleep cycle. The authors state that unlike benzodiazepines and SSRIs, cannabidiol does not disrupt normal sleep architecture. Studies on patients with insomnia are still lacking.

Does more CBD mean a better effect?

No. In the Zuardi 2017 study, anxiety was reduced only by the 300 mg dose, while 100 mg and 900 mg doses did not differ significantly from placebo. The authors described this as an inverted U-shaped dose-response curve, which challenges the popular advice to gradually increase the dose until effective.

Is CBD addictive?

In a study on abuse potential, orally administered cannabidiol up to 800 mg did not differ from placebo in any measure, while smoked marijuana produced effects typical of abused substances. Attachment to the evening ritual itself remains possible and should be periodically reassessed.

Does CBD help with sleep apnea?

There is no data on this. The study circulating in this context involved dronabinol, a synthetic THC, in seventeen patients and showed a reduction in apnea index by 14.1. Cannabidiol itself has not been tested for apnea, and causal treatment is managed by a sleep disorders clinic.

What is the first-line treatment for insomnia?

Cognitive-behavioral therapy for insomnia. The American College of Physicians guidelines from 2016 recommend it as initial treatment for all adult patients with a strong recommendation. Pharmacotherapy is allowed only after its failure, in a shared decision-making process, with a weak recommendation.

Does CBD interact with medications?

Yes. Cannabidiol affects targets involved in drug metabolism, including CYP3A4 and CYP2C19, and P-glycoprotein, with a high potential for interactions with commonly used drugs. EFSA states clearly that safety cannot be established in people taking medications.

Summary

Cannabidiol may help some people whose insomnia stems from tension and anxiety, but the evidence is weaker than the market suggests. The most cited study had no control group or specified dose, and the only polysomnographic placebo-controlled study showed no effect in healthy people. Additionally, the dose-response relationship is not linear.

On safety, the picture is clearer than on efficacy. A meta-analysis of randomized trials shows a liver signal, strongest where cannabidiol meets other drugs, and EFSA explicitly refuses to assess safety in three groups: people under 25, pregnant and breastfeeding women, and those taking medications. The provisional safe dose is about 2 mg per day for a 70 kg person.

If one thing should remain, let it be the order of actions. For insomnia lasting more than three months, the first step is a doctor visit and cognitive-behavioral therapy, not a preparation. For snoring with breathing pauses, the first step is a sleep study. Cannabidiol enters this picture at best as a supportive element after discussion with someone who knows your medications.

It is also worth knowing what is consciously missing from this text. There is no dosing table for sleep because no effective dose has been established and the best-designed study shows a nonmonotonic relationship. There is no percentage of users losing effect after weeks because no study measured this. There is no comparison of bioavailability of different preparation forms because except for inhalation, none has been calculated in humans. The gaps in this article correspond to gaps in the literature and were easier to leave empty than fill with secondhand numbers.

This article is informational and educational and does not constitute medical advice. Before starting cannabis or CBD for therapeutic purposes, consult a doctor, especially if you take other medications, are pregnant, or breastfeeding.

Author: Michał Waluk · Published: 2026-05-06 · Updated: 2026-08-10