
CBD for stress - how much to take, when, and for how long? A practical guide 2026
CBD and stress without promises: what has been measured in anxiety studies, what dose EFSA considered safe, and why the numbers from experiments are not recommendations.
The question of cannabidiol and stress comes up in every conversation about natural support, and the answers circulating online are surprisingly confident: a specific number of milligrams, a ready weekly plan, a table according to body weight. The scientific literature does not contain such things. However, it does include several well-described experiments in which people were given specific amounts of substances under specific conditions, as well as a fresh safety assessment that clearly states for whom safe amounts cannot be determined today. This text shows both, without translating laboratory results into recommendations for the reader. You will see who was studied, how many participants there were, how long the observation lasted, and what exactly was measured, as well as what is missing in these works, although it is often attributed to them.
KEY INFORMATION
- The milligram numbers provided below describe specific studies: who participated, how many participants there were, and how much they were given. They are not a guideline for how much the reader should take.
- In 2026, EFSA derived a provisional safe dose of 0.0275 mg per kilogram of body weight per day, which is about 2 mg daily for a person weighing 70 kg (EFSA 2026).
- The safety of CBD cannot be established today for individuals under 25 years of age, pregnant and breastfeeding women, and those taking medications.
- In the study with simulated public speaking, only the 300 mg dose reduced anxiety; 150 mg and 600 mg did not differ from placebo (Linares 2019).
- In cases of severe anxiety, panic attacks, or low mood, the first step is to talk to a doctor, not to buy a supplement.
What is stress and what does the endocannabinoid system have to do with it?
Stress is the body's reaction to a stimulus requiring adaptation, primarily driven by the hypothalamic-pituitary-adrenal axis and the sympathetic nervous system. Its hormonal marker is cortisol, which has a clear circadian rhythm: it rises in the morning and falls throughout the day. This rhythm will be important in one of the studies described below.
The endocannabinoid system is a network of receptors and compounds produced by the body, present among other things in brain structures responsible for emotions and memory. It participates in extinguishing the stress response, that is, bringing the body back to its baseline state after the stimulus has ceased. This is why it has become a subject of interest for researchers studying anxiety.
Cannabidiol does not strongly bind to the main receptors of this system and does not act as a psychoactive substance. The described mechanisms of its action are indirect and still a subject of debate; the literature points to an influence on serotonin transmission and the prolongation of the action of the body's own compounds. This is at the level of mechanistic hypotheses, not established facts, and should be read as such.
Caution has a practical basis here. Tension, sleep problems, and irritability usually have more than one cause, some of which have nothing to do with the endocannabinoid system: lack of sleep, untreated thyroid disease, chronic pain, low mood. Before concluding that you are lacking a supplement, it is worth ruling out what can be examined. One of the less obvious leads is the gut, which we discuss in the text about how gut bacteria affect mood and stress. how gut bacteria affect mood and stress.
What does CBD do in the brain under stress?
The most concrete answers come from two small imaging and hormonal studies from years ago. Both are cited in guides more often than they are read, and both say something different than what is attributed to them. It is worth knowing them in their original form, as they well illustrate the scale of the knowledge available today.
In the imaging study, ten previously untreated patients with generalized social anxiety participated. Cerebral blood flow was measured at rest using SPECT, twice, after a single oral dose of 400 mg of cannabidiol or placebo. After the active substance, significantly lower subjective anxiety and lower uptake of the tracer in the left parahippocampal gyrus, hippocampus, and inferior temporal gyrus were recorded, and higher in the right cingulate gyrus (Crippa i wsp. 2011).
Two things in this description are important. It was single-photon emission tomography, not functional resonance, and it concerned perihippocampal structures, not the amygdala, which appears most often in popular summaries. The summary of this work also does not provide any percentage decrease in activity, so circulating numbers of this kind have no basis in it.
The second study concerned hormones. Eleven healthy volunteers received placebo or cannabidiol in amounts of 300 mg (seven people) or 600 mg (four people) in the morning. Cortisol levels in sessions with placebo significantly decreased, according to the normal circadian rhythm, and after cannabidiol, this decline was significantly weakened. The substance also had a calming effect on self-assessment scales (Zuardi i wsp. 1993). The authors conclude that cannabidiol interferes with cortisol secretion, not that it lowers it.
What doses were administered in studies on anxiety?
From 150 to 600 mg at once in experiments on anxiety induced by situations, and about 25 mg daily in the only larger observational series from a psychiatric clinic. These two worlds are separated by everything: the number of participants, duration, measurement methods, and whether the subjects were simultaneously taking other medications.
The following summary organizes five studies most frequently cited on this topic. Read them as a description of what was done in the laboratory, not as a list of doses to try. None of these studies were designed to determine a safe amount for a person purchasing a product in a store.
| Study | Who and how many participants | Jak podano | Co wykazano |
|---|---|---|---|
| Linares 2019 | 57 healthy men | A single oral dose of 150 mg, 300 mg, 600 mg, or placebo, before a simulated public speaking event. | Only the 300 mg dose reduced anxiety. |
| Bergamaschi 2011 | 24 untreated patients with social phobia plus 12 healthy individuals. | A single dose of 600 mg or placebo, one and a half hours before the test. | Reduced anxiety and discomfort during the presentation, results similar to healthy individuals. |
| Crippa 2011 | 10 untreated patients with social anxiety. | Jednorazowo 400 mg albo placebo, obrazowanie SPECT w spoczynku | Reduced subjective anxiety and changes in blood flow in the structures of the limbic system. |
| Zuardi 1993 | 11 healthy volunteers. | In the morning, 300 mg (7 people) or 600 mg (4 people). | Diminished normal morning cortisol drop and calming effect. |
| Shannon 2019 | 72 adult patients from a psychiatric clinic. | Almost all took 25 mg daily in capsules, alongside their existing treatment. | Decrease in anxiety scores in the first month for 57 individuals. |
The second row is often summarized most freely. Twenty-four previously untreated patients with social phobia were randomly assigned to 600 mg of cannabidiol or placebo, with twelve people in each group, and an additional twelve healthy individuals underwent the same test without any preparation. After the active substance, anxiety, discomfort, and cognitive disturbances during the speech were lower, and the results approached those of the healthy group (Bergamaschi i wsp. 2011). The authors themselves describe their study as preliminary.
Note the last row. Shannon's series is a retrospective review of medical records, not a study with a control group, and most patients continued their existing psychiatric medications. Attributing improvement solely to cannabidiol is not possible in this setup, and the authors themselves state that controlled studies are needed.
Why can't these results be transferred to everyday life?
Because each of these studies answers a narrower question than what a person seeking help for chronic tension asks. The difference is not that the studies are weak, but that they measured something different. Three limitations regularly recur in them.
The first is participant selection. In the comparison of three doses, only healthy men participated, while in the other two experiments, individuals with diagnosed social anxiety, previously untreated, were included. These groups were selected for measurement purity, not for reflecting the population in which the product is used. There were no women at all in the first of these studies.
The second is the number of subjects and duration. The groups consisted of four to fifteen individuals, and the observation covered one afternoon in the laboratory. With such small numbers, the result of a single study is a signal to check, not a conclusion, which is why there is talk of the need for replication.
The third is the stimulus itself. A simulated public speaking event is a short, predictable, and deliberately induced stress that subsides with the end of the procedure. Chronic tension related to work, caring for a loved one, or a difficult financial situation operates differently and lasts for months. Transferring results from one to the other is an assumption of the researcher reading the work, not a conclusion from that work.
Why does a larger dose not mean a stronger effect?
Because in the only study that deliberately compared several doses in people, the middle dose worked, not the largest. Fifty-seven healthy men were randomly assigned to four groups: 150 mg, 300 mg, 600 mg of cannabidiol, or placebo, in a double-blind procedure, before a simulated public speaking test.
Compared to placebo, anxiety during the speech was significantly reduced only by the 300 mg dose. The groups receiving 150 mg and 600 mg did not differ from placebo in self-reported mood scales. The authors described this as a dose-response curve in the shape of an inverted U and considered it a confirmation of earlier observations from animal studies (Linares i wsp. 2019).
The conclusion the authors draw from this is cautious and worth quoting in full: optimal therapeutic amounts still need to be rigorously determined for the study results to be applicable to clinical practice. This sentence is from the abstract, not an interpretation. In other words, the researchers state that there is no transfer yet.
For the reader, this means two things. First, increasing the amount on your own is not a strategy supported by results, as in the only such comparison, the largest dose performed worse than the average. Second, the result obtained in healthy men before one presentation says nothing about daily use over weeks, which this study did not cover at all.
What is known today about the safe amount of CBD?
The latest answer comes from the EFSA update on the safety of cannabidiol as a novel food. The panel derived a provisional safe dose of 0.0275 mg per kilogram of body weight per day, with an uncertainty factor of 400, which translates to about 2 mg daily for a person weighing 70 kg (EFSA 2026).
This value has a narrowly defined range. It applies only to supplements with a purity of cannabidiol of at least 98 percent, without nanoparticles, produced through a safe manufacturing process, and excluding genotoxicity. It does not apply to other forms or products with unknown compositions.
The most important sentence of this assessment does not contain a number. The panel states that based on all available safety data, it is impossible to determine the safety of cannabidiol for individuals under 25 years of age, pregnant women, breastfeeding mothers, and individuals taking medications simultaneously. A reader seeking support for chronic tension often belongs to one of these three groups.
The scale of discrepancy between these two milligrams and hundreds of milligrams from experiments can be misleading, so it is worth naming it directly. The experiment checks what the substance does in one situation with a narrowly selected group under supervision. The safety assessment asks what amount can be considered safe for daily consumption by any person in the population, and therefore applies a large uncertainty factor. These are two different questions, not two contradictory answers.
How long did the observations last in the studies?
From one session to three months. Experiments on situational anxiety measured the effect of a single dose given in one afternoon. The only larger observational series tracked patients over subsequent monthly visits, but without a control group and without random assignment.
In this series, the documentation of 103 patients was analyzed, of which 72 adults were included in the study: 47 primarily reported anxiety, and 25 primarily reported poor sleep. Anxiety scores were measured using the Hamilton scale, and sleep quality was assessed with the PSQI questionnaire at each monthly visit (Shannon i wsp. 2019).
The results are mixed and have been described as such. In the first month, anxiety scores decreased in 57 individuals, or 79.2 percent of those studied, and remained lower in subsequent months. Sleep scores improved in 48 individuals, or 66.7 percent, but varied over time, and in some patients, symptoms worsened. In the anxiety group, the average Hamilton scale score was 23.87 at the start and 16.36 after three months.
The authors emphasize the limitations of their own work: it is a review of documentation, not a controlled study, patients remained under psychiatric care, and most continued their medications. The conclusion they formulate is that cannabidiol may be beneficial in anxiety-related disorders and that controlled clinical trials are needed. This is the full strength of the evidence available today.
What do the scales used in these studies actually measure?
The percentages cited in guides come from specific measurement tools, not from a general impression of improvement. It is important to know what these tools count, as it changes the way results are interpreted. Without this, the statement about improvement in seven out of ten individuals sounds stronger than it deserves.
The Hamilton anxiety scale has fourteen questions and a scoring range from 0 to 56. A score below 17 points corresponds to mild anxiety, while above 25 points indicates severe anxiety. In the series from the psychiatric clinic, the average score in the anxiety group was 23.87 points at the start, placing it between these thresholds, and after three months, it was 16.36 points.
Sleep quality was measured using the PSQI questionnaire, consisting of nineteen items rated from 0 to 3, with a total range from 0 to 21 points. A higher number indicates more problems, and a score of 5 points or higher classifies the subject as a poor sleeper. Both scales are self-reporting and clinical tools, not physiological measurements.
In experiments on public speaking, different tools were used: a visual analog mood scale and a negative self-thought questionnaire, supplemented by measurements of blood pressure, heart rate, and skin conductance. These are also subjective assessments, supplemented by a few physiological parameters. None of these studies measured anything that could be called an objective level of stress.
Does tolerance develop to CBD, and do you need to take breaks?
It is unknown, and that is a fair answer. There is no study that has checked in humans whether repeated intake of cannabidiol weakens its effects over time, nor one that compares continuous use with intermittent use. Claims about the percentage of individuals losing response after a certain number of weeks have no source in the literature.
The same applies to the trendy explanation of increasing the number of receptors in response to the presence of the substance. This is a hypothesis described for other compounds acting on this system, transferred to cannabidiol by analogy, not based on measurement. A mechanistic explanation without experimental results is a story, not evidence.
What can be said sensibly? That the subjective impression of the effects of any substance with a mild profile changes with circumstances: with the season, workload, amount of sleep, and what we expect. Distinguishing a change in the body's reaction from a change in life situation requires recording, not memory, because memory in such matters is unreliable.
If you are looking for signals by which the sense of a break is practically assessed, we have described them separately in the text about when it's time for a tolerance break. In any doubts regarding health, make the decision to continue or discontinue anything with your doctor.
What will CBD not do?
It will not replace the treatment of anxiety disorders or depression, it does not act like a sedative taken on demand, and it does not remove the causes of tension that lie outside the body. Cannabidiol is not registered in Poland as a medication for use in stress, and the products available for sale are not medicinal products.
It is also worth distinguishing between two different states that are commonly referred to by the same name. Tension accompanying a difficult period is an adaptive response. An anxiety disorder is a medical diagnosis, with specific criteria and effective treatment methods, primarily psychotherapy and pharmacotherapy. Replacing the latter with a supplement means postponing the help that works.
Separate are situations requiring urgent response: panic attacks, suicidal thoughts, insomnia lasting for weeks, unexplained somatic symptoms. These are not situations where a reasonable first step is to purchase anything. These are situations where you call a doctor.
Finally, caution regarding your own expectations. The studies described above measured changes in self-assessment scales and clinical scores, usually moderate, in small groups of individuals. No one in these works described the disappearance of anxiety or a lasting change in reactivity to stress. Promises of this kind come from marketing materials, not from the literature.
It will also not do one thing that is rarely mentioned: it will not become safer just because it is plant-based. The EFSA assessment from 2026 indicates consistent liver toxicity in animal studies and potential hepatotoxicity in humans when used concurrently with medications. Natural origin is not a toxicological category, and in this case, that difference can be the most misleading for the reader.
Where to start before reaching for a supplement?
From two things that cannot be bought: from talking to a doctor if you are taking any medications, and from naming what exactly you want to change. The EFSA assessment states clearly that for individuals taking medications, the safety of cannabidiol cannot be determined, so this conversation is not a formality.
The second thing is a reference point. Without a recorded baseline state after a few weeks, you will not distinguish improvement from a good week, or worsening from a worse mood on Monday. A short daily note is sufficient: how you slept, how you assess your tension, what was happening at work and at home during that time. We described a template for such a record in the text about how to keep a cannabinoid journal.
The third thing concerns the product itself, if after those two you still want to try it. Only the certificate of analysis for a specific batch is verifiable, with a number matching the packaging, the date of testing, and the indicated cannabinoid content. Everything else in the offer is a seller's declaration.
The fourth thing is the basics that are easy to forget when looking for a preparation: regular sleep, exercise, and limiting stimulants. They do not sound attractive, and that is precisely why they are often overlooked. A supplement added to a disrupted daily rhythm usually has nothing to support.
The fifth thing is a predetermined end to the trial. Decide before you start how you will know that it is not worth continuing, and when you will make that assessment. Without such a determination, the trial has no end, and its extension is decided by habit and money already spent, rather than observation. In the only longer series described above, the assessment took place during monthly medical visits, and this is a reasonable reference point even when no one is guiding you by the hand.
What are the risks and drug interactions?
The most serious risk is the impact on the metabolism of other drugs. A study on the inhibition of liver cytochrome P450 enzymes by cannabinoids and their metabolites showed that cannabidiol competitively inhibits CYP3A4, CYP2B6, CYP2C9, CYP2D6, and CYP2E1, while acting in a mixed manner on CYP1A2 (Nasrin i wsp. 2021).
These enzymes are responsible for the metabolism of many drugs used chronically, including those for which even a slight change in blood concentration has clinical significance. Therefore, in any prescription treatment, the decision is made by a doctor or pharmacist, not by a seller or the author of a guide. The time interval between a drug and a supplement does not eliminate this problem, as enzyme inhibition lasts longer than the presence of the substance in the digestive tract.
The second group of risks concerns the liver. In the EFSA assessment, animal studies showed consistent liver toxicity, with liver mass and histopathological changes as sensitive endpoints, and human studies indicated a potential hepatotoxicity, especially with concurrent use of medications. Higher amounts have also been associated with gastrointestinal complaints.
The panel also noted the penetration of cannabidiol through the placenta and accumulation in the body, neurodevelopmental effects from long-term prenatal exposure that are sex-dependent, as well as hormonal disturbances, including altered thyroid hormone levels. Immunotoxicity has not been studied at all, which the panel explicitly states. The lack of study is not a reassuring outcome.
Is CBD legal in Poland?
Cannabidiol itself does not appear on controlled substance lists, and products from industrial hemp remain legally available. The legality is determined by the raw material and the content of tetrahydrocannabinol in the plant material from which the product was made, not by the brand name or declared purpose.
The law defines industrial hemp as plants in which the sum of delta-9-THC and tetrahydrocannabinolic acid content in flowering or fruiting tops, from which resin has not been removed, does not exceed 0.3 percent when calculated on a dry weight basis, with the sum rounded to one decimal place. The basis is Article 4 point 5 of the Act of July 29, 2005 on counteracting drug addiction (Journal of Laws 2023, item 1939), as amended. the Act of March 24, 2022 (Dz.U. 2022 poz. 763)..
It is worth separating the two regulations with the same numerical value. The national threshold corresponds to the EU threshold, but it does not imply: the EU's 0.3 percent for varieties eligible for support under the Common Agricultural Policy was introduced by Regulation 2021/2115, effective from January 1, 2023. The previous EU threshold was 0.2 percent.
A separate issue is the food status. Cannabidiol remains subject to the novel food procedure in the European Union, and a health notification is not an authorization, while the EFSA position from 2026 is a safety assessment with conditions, not a clearance for food circulation. This also implies a ban on attributing therapeutic effects to such products and using unauthorized health claims.
This last principle explains the apparent inconsistency seen in every store. A product description cannot promise calming or improved sleep, as such claims for cannabidiol have not been authorized, yet articles online state this outright. Therefore, if the product card promises effects on stress, it is information from the seller, not about the product. A reliable offer, instead of a promise, shows the composition and the results of batch testing.
Frequently Asked Questions
How much CBD should I take for stress?
This article does not resolve and cannot resolve that. The numbers from studies describe specific experiments, not recommendations for the reader, and EFSA stated that for individuals taking medications, pregnant, breastfeeding, and under 25 years of age, a safe amount cannot be determined today. Decide on the amount with your doctor.
What amount of CBD has been considered safe?
In 2026, EFSA derived a provisional safe dose of 0.0275 mg per kilogram of body weight per day, which is about 2 mg daily for a person weighing 70 kg. This value applies only to supplements with a purity of at least 98 percent, without nanoparticles. No amount has been established for other forms.
Does a higher dose work stronger on anxiety?
In the only study that compared several amounts in humans, no. Fifty-seven healthy men were given a single dose of 150 mg, 300 mg, 600 mg, or placebo before a simulated public speaking event. Only the 300 mg dose reduced anxiety, while the other two did not differ from placebo.
Does CBD lower cortisol?
Not in the usual way it is summarized. In a 1993 study involving eleven healthy volunteers, morning cortisol levels decreased according to the circadian rhythm during placebo sessions, and after cannabidiol, this decline was significantly weakened. The authors speak of interference with cortisol secretion, not its reduction.
How long did the studies on CBD and anxiety last?
From one laboratory session to three months. The longest observation is a retrospective review of the documentation of 72 patients from a psychiatric clinic, without a control group, in which most individuals continued their previous medications. The authors themselves state that controlled studies are needed.
Czy do CBD powstaje tolerancja?
It is unknown. There is no study that has examined the diminishing effect of cannabidiol over time in humans or compared continuous use with intermittent use. The circulating percentages of individuals losing response after a certain number of weeks have no source in the literature, and receptor explanations are transferred by analogy.
Can CBD be combined with anxiolytics?
Not without consultation. In a study on inhibiting liver enzymes, cannabidiol competitively inhibited CYP3A4, CYP2B6, CYP2C9, CYP2D6, and CYP2E1, which are enzymes responsible for the metabolism of many drugs. EFSA stated that for individuals taking medications, the safety of cannabidiol cannot be determined.
Is CBD legal in Poland?
Yes, as long as the product comes from hemp. The law defines them by the sum of delta-9-THC and tetrahydrocannabinolic acid not exceeding 0.3 percent of dry weight, rounded to one decimal place. Cannabidiol is not listed among controlled substances, but it is also not authorized as a novel food.
Co z tego wynika dla czytelnika
The available evidence today is much more modest than guides suggest. A few small experiments showed a reduction in anxiety after a single large dose in public speaking situations, one observational series without a control group described improvement in patients taking medications simultaneously, and the safety assessment from 2026 indicated a quantity lower than those by two orders of magnitude.
From such a set, it is not possible to honestly derive a plan for the reader, and anyone who does so is not relying on the literature. However, three practical conclusions can be drawn: increasing the amount on your own has no support in the results, the decision regarding any medications lies with the doctor, and in cases of heightened anxiety, the first step is diagnosis, not purchase.
If you still want to try, treat it as an experiment on one person, conducted with record-keeping and a reasonable time horizon. Then at least you will learn something about yourself, not about the effectiveness of advertising.
This article is for informational and educational purposes and does not constitute medical advice. Before starting to use cannabis or CBD for therapeutic purposes, consult a doctor, especially if you are taking other medications, are pregnant, or breastfeeding.
Author: Michał Waluk · Opublikowano: 2026-05-11 · Aktualizacja: 2026-08-10







