The best trial in this area found a real, sizeable improvement in insomnia severity. In the same patients, on the same nights, the sleep laboratory found almost no difference in their actual sleep. Both results deserve to be taken seriously, but neither is the main reason for caution. The main reason is what happens when you stop.
This is general information, not medical advice, and not a recommendation to use cannabis for sleep. Chronic insomnia is usually a symptom of something else: pain, anxiety, depression, sleep apnoea, a medication or a schedule. Treating the symptom without finding the cause is how people end up years into a problem that had an answer. In Thailand, cannabis flower is a controlled herb. It is dispensed only on prescription form PT 33 from an authorised practitioner, and recreational use is not permitted.
1. What the Older Literature Says
The earliest controlled hypnotic study, Cousens and DiMascio (1973), gave THC at 10, 20 and 30 mg to people with insomnia. People fell asleep faster, and side effects increased with the dose. The next-day “hangover” at 30 mg was judged severe enough to rule that dose out as a sleep medicine. The popular account still rests largely on that picture: faster sleep onset and dose-related grogginess the next day. REM suppression usually gets only a passing mention, and it deserves more attention than that (section 4).
2. The Best Trial
Walsh and colleagues, SLEEP, 2021. A randomised, double-blind, placebo-controlled crossover trial of a cannabinoid medicine (ZTL-101) in chronic insomnia.
Participants. 24 randomised and 23 completed and analysed. Mean age was about 53, and 83% were female. Each person took the sublingual product and a placebo for two weeks each, with a one-week washout in between.
Product. THC 20 mg/mL, CBN 2 mg/mL and CBD 1 mg/mL, with terpenes, in oil. The protocol was 0.5 mL about an hour before bed. From night 4, participants could increase to 1.0 mL with a physician’s approval and could reduce the dose at any time. This describes the trial protocol and is not a recommended dose.
Results.
| Measure | Result versus placebo |
|---|---|
| Insomnia Severity Index | −5.07 points (p = 0.0001, d = 0.94) |
| Total sleep time (actigraphy) | +33.4 minutes (p < 0.001) |
| Wake after sleep onset (actigraphy) | −10.2 minutes (p = 0.002) |
| Sleep efficiency (actigraphy) | +2.9% (p ≤ 0.005) |
| Polysomnography (sleep stages and architecture) | No significant difference |
All adverse effects were mild. Among participants taking the active product, 33% had dry mouth, 25% dizziness and 17% headache. Almost all of these effects (97.5%) resolved overnight or soon after waking. There were no serious adverse events.
By the standards of this field, that is a genuinely positive trial. It showed a large effect on the primary outcome, backed by objective actigraphy.
3. The Two Problems Inside It
The laboratory could not see what the questionnaire measured. Actigraphy (a wrist movement monitor) showed clear improvements. Polysomnography, the full sleep-laboratory measurement of sleep stages and architecture, showed no significant difference between drug and placebo. The two instruments measure different things: actigraphy infers sleep from stillness, while polysomnography records the brain’s electrical activity during sleep. A drug that makes you lie still has not necessarily changed your sleep.
The same pattern appears in other cannabinoid trials: the patient-reported outcome improves and the objective measure does not follow. That does not mean patients are wrong. It means the claim being supported is narrower than it sounds.
The blinding failed. The authors say so directly: the treatment effects were noticeable enough that effective blinding could not be maintained. Every participant who answered correctly guessed when they were on the active product. When the primary outcome is subjective, participants knowing their arm is a serious limitation. It also matters because of a 2022 meta-analysis of cannabis pain trials in JAMA Network Open (Gedin et al.), which found that placebo alone produced a moderate-to-large reduction in pain scores.
Add the sample size of 23 people over two weeks, and the honest description is a promising signal that needs a bigger and longer trial. It is not a settled result.
4. The REM Problem
Sleep runs in cycles of roughly 90 minutes. Each cycle passes through lighter NREM stages (N1 and N2) and deep slow-wave sleep (N3), which is concentrated in the first half of the night. Then comes REM sleep, which makes up roughly a fifth to a quarter of adult sleep and lengthens towards morning. REM is linked to vivid dreaming and is thought to play a part in emotional memory processing. A sleep aid therefore has to be judged on what it does to this structure, not only on how fast you fall asleep.
Studies generally report reduced REM with THC, with REM rebound after stopping; the long-term significance is not well established. The effect appears to depend on dose and on tolerance, and the human data are partly mixed (Babson, Sottile & Morabito, 2017). Some conventional hypnotics and antidepressants also reduce REM, and that is one of the criticisms made of them.
Two consequences follow. Someone who dislikes their dreams, such as a patient with PTSD nightmares, may experience less REM as the benefit. And when use stops, REM tends to rebound, with unusually vivid, intense dreams for a while afterwards. That leads to the part that matters most.
5. The Trap
Stopping abruptly after regular use produces a recognised withdrawal syndrome, and trouble sleeping and strange or vivid dreams are among its core features. Bahji and colleagues (2020) ran a meta-analysis of 47 studies with 23,518 participants. They found cannabis withdrawal syndrome in 47% of people with regular or dependent use. The rate varied widely by setting: about 17% in general-population samples, 54% in outpatients and 87% in inpatients. Clinical descriptions say symptoms usually begin within 1–3 days, peak in the first week and settle over 1–2 weeks. That timing comes from clinical descriptions and was not established by the meta-analysis.
Sleep laboratory data point the same way. Bolla and colleagues (2008) recorded heavy users in the first nights after stopping. Compared with non-users, they slept less, had less slow-wave sleep and lower sleep efficiency, and took longer to fall asleep. Because the study compared two groups at one point in time, it cannot show how much of this difference existed before the users stopped.
Now put that beside a person using cannabis for insomnia. After some months they try to stop, and their sleep becomes markedly worse than before. The natural conclusion is this proves I need it, but that can’t be tested from the inside, because withdrawal insomnia and the original insomnia feel the same.
That is the trap: after a period of regular use, you can no longer tell whether you are treating the insomnia you started with or preventing the insomnia that stopping now causes. This is not unique to cannabis. Every hypnotic has the same problem, and it is one reason sleep medicine moved away from long-term hypnotic use. It applies here for the same reasons.
The practical implication is not “never.” It is that an exit plan belongs in the decision from the beginning. That means a defined duration, a taper agreed with the prescriber rather than abrupt cessation, and an expectation, set in advance, that the first week or two after stopping will be worse than baseline. Those bad nights are not evidence of anything.
6. What About CBN and CBD?
Products marketed for sleep often feature CBN, which forms as THC breaks down when cannabis ages. There are now randomised trials of CBN, and the results are modest:
- Bonn-Miller et al. (2024) was an industry-linked trial of 321 adults who rated their sleep as poor. Taking 20 mg of CBN nightly for a week did not significantly improve the primary outcome, sleep quality, compared with placebo. It did reduce night-time awakenings and overall sleep disturbance, and adding CBD brought no extra benefit.
- Lavender et al. (2026) was an Australian single-night crossover trial in 20 adults with diagnosed insomnia. Neither 30 mg nor 300 mg of CBN changed the primary polysomnography outcome, wake after sleep onset. The 300 mg dose shortened the time to fall asleep by about 7 minutes and improved how well people felt they had slept. The authors concluded that the trial does not provide direct evidence that CBN works for insomnia in routine care.
For CBD on its own, a single 300 mg dose had no acute effect on the sleep of healthy volunteers (Linares et al., 2018). The improvement often quoted from a 72-patient case series (Shannon et al., 2019) came from an uncontrolled clinic sample treated mainly for anxiety. Neither CBN nor CBD has the evidence to be called a sleep treatment.
7. What Has the Best Evidence for Chronic Insomnia
Cognitive behavioural therapy for insomnia (CBT-I). The American College of Physicians recommends it as the initial treatment for chronic insomnia disorder in adults, ahead of medication (Qaseem et al., 2016).
CBT-I is not relaxation advice. It is a structured programme, usually delivered over a few weeks and increasingly available online. It combines sleep restriction, stimulus control and cognitive work on sleep-related thinking, with sleep hygiene as a minor part. It takes more effort than taking something at bedtime. Unlike any hypnotic, it produces improvements that last after treatment ends, with no withdrawal.
Any honest article about a sleep aid has to name it, because most people reading about cannabis for sleep have never been offered it.
8. Where a Cannabinoid Might Reasonably Fit
Under a prescriber’s supervision, and within these limits:
- Short-term, with a defined duration and a tapering exit, rather than open-ended nightly use.
- For insomnia secondary to something a cannabinoid is already prescribed for, such as pain, nausea or spasticity. Here better sleep is a side benefit of treating the main problem.
- Not as a substitute for finding the cause. Sleep apnoea, depression, thyroid disease, restless legs and medication effects can all show up as insomnia, and each has its own treatment.
- Not where next-day impairment matters. Higher doses cause grogginess the next day, so the usual driving and machinery cautions apply the following morning, not only at night.
- With particular care in older patients. They are more sensitive to sedation and to the drop in blood pressure on standing that causes falls. A night-time dose followed by a trip to the bathroom at 3 a.m. is a specific and predictable risk.
9. In Thailand
Cannabis flower is a controlled herb under the Notification of the Ministry of Public Health Re: Controlled Herbs (Cannabis), B.E. 2568, in force since late June 2025. It can be dispensed only against a prescription on form PT 33 from an authorised practitioner. Each prescription covers no more than 30 days and is single use (no refills). Flower may not be sold or advertised online, and recreational use is not lawful.
Licensing was tightened further by two Ministerial Regulations issued in 2026 (B.E. 2569):
- One limits cannabis and hemp extracts above 0.2% THC to medical, research and other permitted purposes.
- The other raises the requirements for licensed flower sellers.
Authorities also reported nationwide inspections of cannabis shops in mid-2026. The Cannabis Control Act is still a draft: the Cabinet approved it on 22 September 2026 and sent it to the House.
These rules are still changing, so check current requirements with your prescriber or the Department of Thai Traditional and Alternative Medicine. A time-limited prescription has one advantage for a sleep aid: long-term nightly use becomes visible and has to be reviewed regularly, instead of quietly accumulating.
If you have used a cannabinoid for sleep under a prescription for more than a few months, the most useful thing you can post is what happened when you stopped. How long did the bad nights last, and did your sleep eventually settle above or below where it started? This area has almost no data on that question, and no trial has followed anyone long enough to answer it. Share only experience under a lawful prescription; do not post sources, prices or recreational use.
References
- Bahji A, Stephenson C, Tyo R, Hawken ER, Seitz DP. Prevalence of cannabis withdrawal symptoms among people with regular or dependent use of cannabinoids: a systematic review and meta-analysis. JAMA Netw Open. 2020;3(4):e202370. doi:10.1001/jamanetworkopen.2020.2370
- Babson KA, Sottile J, Morabito D. Cannabis, cannabinoids, and sleep: a review of the literature. Curr Psychiatry Rep. 2017. doi:10.1007/s11920-017-0775-9
- Bolla KI, et al. Sleep disturbance in heavy marijuana users. Sleep. 2008;31(6):901–908. PMID 18548836
- Bonn-Miller MO, et al. A double-blind, randomized, placebo-controlled study of the safety and effects of CBN with and without CBD on sleep quality. Exp Clin Psychopharmacol. 2024;32(3):277–284. doi:10.1037/pha0000682
- Cousens K, DiMascio A. (−)Δ9 THC as an hypnotic: an experimental study of three dose levels. Psychopharmacologia. 1973;33:355–364. doi:10.1007/BF00437513
- Gedin F, et al. Placebo response and media attention in randomized clinical trials assessing cannabis-based therapies for pain: a systematic review and meta-analysis. JAMA Netw Open. 2022;5(11):e2243848. doi:10.1001/jamanetworkopen.2022.43848
- Lavender IG, et al. Cannabinol for acute treatment of insomnia disorder in a randomized placebo-controlled crossover trial. J Sleep Res. 2026;35(4):e70284. doi:10.1111/jsr.70284
- Linares IM, et al. No acute effects of cannabidiol on the sleep-wake cycle of healthy subjects. Front Pharmacol. 2018;9:315. doi:10.3389/fphar.2018.00315
- Qaseem A, et al. Management of chronic insomnia disorder in adults: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2016;165(2):125–133. doi:10.7326/M15-2175
- Shannon S, et al. Cannabidiol in anxiety and sleep: a large case series. Perm J. 2019;23:18-041. doi:10.7812/TPP/18-041
- Walsh JH, Maddison KJ, Rankin T, et al. Treating insomnia symptoms with medicinal cannabis: a randomized, crossover trial of the efficacy of a cannabinoid medicine compared with placebo. Sleep. 2021;44(11):zsab149. doi:10.1093/sleep/zsab149
General information, not medical advice. Persistent insomnia deserves a diagnosis, not just a remedy.