Weed and Sleep: What THC Actually Does to Your REM

THC suppresses REM sleep. Controlled EEG studies show it reduces rapid eye movement activity and shortens REM itself, and the effect is largest in the first nights of use. Tolerance builds quickly, so regular users often keep the habit long after the sedation has faded. Stopping produces rebound: more REM, unusually vivid dreams, and several days of fragmented sleep. Population data links recent use to both short and long sleep. Falling asleep faster and sleeping well are different outcomes, and THC mostly delivers the first one.

Cannabis is the most common self-prescribed sleep aid in the country, and almost nobody using it that way has looked at what the sleep lab data shows. The research here is older and thinner than you'd expect for something this widely used, but the part about REM is one of the more consistent findings in the whole substance-and-sleep literature. This is a description of physiology, not a verdict on anyone's choices. What follows is what polysomnography and population data actually show, including the places where the evidence runs out.

1. What THC does to REM sleep

The clearest human data comes from a 1975 EEG study that dosed experienced cannabis users with THC and recorded them across baseline, low dose, high dose and withdrawal nights. THC significantly reduced eye movement activity during REM, and to a lesser degree shortened REM duration itself. The authors described these as the most pronounced effects on brain electrical activity any drug had shown at that point.

Citation: Feinberg I, Jones R, Walker JM, et al. Effects of high dosage delta-9-tetrahydrocannabinol on sleep patterns in man. Clinical Pharmacology and Therapeutics, 1975. PubMed: 164314

Worth naming the weakness right away. That study ran seven subjects total, four of them through the full protocol. Small sample, 1970s methodology, pharmaceutical-grade THC rather than what anyone buys today. The finding has held up across later reviews, but it rests on a narrow base.

REM is where most vivid dreaming happens, and it's the stage associated with emotional processing and certain kinds of memory consolidation. Cutting it isn't neutral, though the functional consequences of a few weeks of reduced REM in a healthy adult are genuinely not well characterized.

Citation: Angarita GA, Emadi N, Hodges S, Morgan PT. Sleep abnormalities associated with alcohol, cannabis, cocaine, and opiate use: a comprehensive review. Addiction Science and Clinical Practice, 2016. PubMed: 27117064

2. Why the first few nights feel like it works

Because they do work, in the narrow sense people mean. Reviews of the human literature find that THC tends to shorten sleep latency, which is the time it takes to fall asleep. If your complaint is lying awake for forty minutes with a busy head, THC addresses that specific problem on night one.

The 2017 review in Current Psychiatry Reports puts it plainly: THC may decrease sleep latency but could impair sleep quality with long-term use. Both halves of that sentence are load-bearing, and most people only ever experience the first half.

Citation: Babson KA, Sottile J, Morabito D. Cannabis, Cannabinoids, and Sleep: a Review of the Literature. Current Psychiatry Reports, 2017. PubMed: 28349316

There's a second reason the first nights feel good. Sedation is easy to notice and REM loss is invisible. You have direct access to how fast you fell asleep. You have no access at all to how your sleep architecture was distributed across the night. The feedback you get is systematically biased toward the one measure THC improves.

3. Tolerance shrinks the effect while the habit stays

This is the part that changes the picture. The sedative effect attenuates with repeated use, and it does so faster than most people notice. The systematic review of 39 human cannabinoid administration studies found mixed results across sleep outcomes, with acute effects differing substantially from effects in chronic users.

Citation: Gates PJ, Albertella L, Copeland J. The effects of cannabinoid administration on sleep: a systematic review of human studies. Sleep Medicine Reviews, 2014. PubMed: 24726015

What you end up with is a routine that no longer produces the effect that created it, plus a withdrawal cost if you skip it. The sleep aid becomes a maintenance requirement. That's the mechanism behind the very common report of "I don't really get high anymore, I just can't sleep without it."

4. What happens when you stop

REM comes back, and it comes back with interest. In the 1975 EEG study, withdrawal produced eye movement activity and REM duration above baseline, with the rebound more pronounced for eye movement than for REM time. That rebound is the physiological explanation for the extremely vivid, often unpleasant dreams people report in the first week after stopping.

The sleep disruption is measurable, not just subjective. A crossover study at Johns Hopkins put 20 daily cannabis users through alternating periods of use and three-day abstinence with polysomnography throughout. During placebo abstinence, participants showed decreased sleep efficiency, decreased total sleep time, shortened REM latency, longer sleep onset and more time in REM compared with their own cannabis-using nights.

Citation: Vandrey R, Smith MT, McCann UD, Budney AJ, Curran EM. Sleep disturbance and the effects of extended-release zolpidem during cannabis withdrawal. Drug and Alcohol Dependence, 2011. PubMed: 21296508

A separate inpatient study recorded 18 heavy users on nights 1, 2, 7, 8 and 13 after abrupt discontinuation. Total sleep time, sleep efficiency and REM amount all declined across abstinence while wake after sleep onset and periodic limb movements increased. Quantity and duration of prior use tracked with more periodic limb movements.

Citation: Bolla KI, Lesage SR, Gamaldo CE, et al. Polysomnogram changes in marijuana users who report sleep disturbances during prior abstinence. Sleep Medicine, 2010. PubMed: 20685163

An earlier study from the same group compared 17 heavy users in early abstinence against 14 drug-free controls and found lower total sleep time and less slow wave sleep on both recording nights, with worse sleep efficiency and shorter REM latency on night two. Notably, the cannabis group didn't improve on the second night the way a control group normally does after an adaptation night. Sample sizes here are small and the design can't separate pre-existing sleep problems from drug effects, which the authors say directly.

Citation: Bolla KI, Lesage SR, Gamaldo CE, et al. Sleep disturbance in heavy marijuana users. Sleep, 2008. PubMed: 18548836

5. What the population data shows

The largest analysis pulled NHANES data from 2005 to 2018, representing roughly 146 million US adults, of whom 14.5 percent reported cannabis use in the past 30 days. After adjustment, recent users were more likely than non-users to report short sleep under six hours (odds ratio 1.34, 95 percent CI 1.12 to 1.59) and also more likely to report long sleep over nine hours (odds ratio 1.56, 95 percent CI 1.25 to 1.96). Heavy users, meaning 20 or more of the past 30 days, sat further out at both extremes.

Citation: Diep C, Tian C, Vachhani K, et al. Recent cannabis use and nightly sleep duration in adults: a population analysis of the NHANES from 2005 to 2018. Regional Anesthesia and Pain Medicine, 2022. PubMed: 34873024

Read that carefully, because it's cross-sectional. It cannot tell you whether cannabis pushed people to the extremes or whether people already sleeping badly reached for cannabis. There's decent evidence for the second direction too. A genetic analysis in the journal Sleep found significant genetic correlation between cannabis use disorder and both short sleep duration and insomnia, which suggests some shared underlying predisposition rather than a clean one-way causal arrow.

Citation: Winiger EA, Ellingson JM, Morrison CL, et al. Sleep deficits and cannabis use behaviors: an analysis of shared genetics. Sleep, 2021. PubMed: 32935850

6. What the evidence does not support

Plenty, and this cuts against the confident claims on both sides. The 2020 systematic review in Sleep Medicine Reviews screened preclinical and clinical work and found 14 preclinical and 12 clinical studies meeting criteria. Its conclusion was that there's insufficient evidence to support routine clinical use of cannabinoid therapies for any sleep disorder, given how little has been published and the moderate-to-high risk of bias in most of what exists.

Citation: Suraev AS, Marshall NS, Vandrey R, et al. Cannabinoid therapies in the management of sleep disorders: a systematic review of preclinical and clinical studies. Sleep Medicine Reviews, 2020. PubMed: 32603954

So the honest summary has three parts. THC reliably suppresses REM, that's well supported. Tolerance and rebound are well documented in daily users. And the claim that cannabis is a good long-term sleep tool has never been demonstrated in a trial designed to test it. Anyone telling you otherwise is ahead of the data.

Where Lunia fits

Lunia Restore is melatonin-free by design, built to support the body's own sleep pathways rather than dosing a hormone. It contains magnesium bisglycinate at 500 mg of the compound, which delivers 90 mg of elemental magnesium, plus L-theanine at 300 mg and apigenin at 50 mg.

On the evidence: in a placebo-controlled trial, magnesium bisglycinate was shown to reduce insomnia severity scores, though that trial used 250 mg elemental magnesium and Lunia delivers 90 mg, roughly 36 percent of the trial dose. We publish that gap rather than hide it. A 2025 meta-analysis of 19 trials found L-theanine improved subjective sleep quality and sleep onset, and research points to a 200 to 450 mg daily range. Apigenin binds to the benzodiazepine site on GABA-A receptors in laboratory studies, and higher dietary apigenin intake has been associated with better sleep quality. There's no randomised trial on the finished Lunia formula. Our evidence is at the ingredient level, at doses and forms we publish in full.

Nothing here is a substitute for a conversation with a clinician, and that's especially true for anyone thinking about changes to a substance they use regularly. Individual results may vary.

Learn more about Lunia Restore

Frequently Asked Questions

Does weed actually help you sleep?

It helps with one part of sleep and hurts another. Human studies consistently show THC shortens the time it takes to fall asleep, which is the effect people notice. The same research shows it suppresses REM sleep, which people can't notice without a sleep lab. Reviews describe THC as decreasing sleep latency while potentially impairing sleep quality over the long term.

How much REM sleep does THC suppress?

There's no single reliable percentage, and anyone quoting one is overstating the data. The 1975 EEG work found THC significantly reduced eye movement activity within REM and shortened REM duration to a lesser extent, across a total of seven subjects. Later reviews confirm the direction of the effect without pinning down a stable magnitude across doses and products.

Why do I get vivid dreams when I stop smoking?

That's REM rebound. When THC is removed, REM activity returns above its baseline level rather than simply back to it. The original EEG study measured exactly this pattern during withdrawal nights, with the rebound stronger for eye movement activity than for total REM time. More REM means more dreaming, and the dreams are often described as unusually intense.

How long does sleep disruption last after stopping cannabis?

Longer than most people expect. The inpatient polysomnography study tracked heavy users on nights 1, 2, 7, 8 and 13 after discontinuation and found total sleep time, sleep efficiency and REM amount still declining across that window. So the disturbance persisted through at least the second week in that sample of 18. Individual timelines vary a great deal, and this is a question worth raising with a clinician.

Is CBD different from THC for sleep?

Pharmacologically, yes, they act differently. Preliminary research suggests CBD may have therapeutic potential for insomnia, and CBD hasn't shown the REM suppression pattern THC has. The evidence base is much smaller though. The 2020 systematic review concluded there's insufficient evidence to support routine clinical use of any cannabinoid therapy for any sleep disorder.

Does daily cannabis use cause long-term sleep problems?

The association is there, the causation isn't settled. NHANES data links recent use to both short and long sleep duration with a dose-response pattern in heavy users. But genetic analysis shows cannabis use disorder shares underlying genetic correlation with insomnia and short sleep, which means some of that association likely runs in the other direction. Both things can be true at once.

Does cannabis affect sleep apnea?

Some early work on synthetic cannabinoids like nabilone and dronabinol suggested short-term benefit for obstructive sleep apnea through effects on serotonin-mediated apneas. That work hasn't matured into anything approaching a standard of care, and the 2020 review lists sleep apnea as an area needing randomised trials rather than one with an answer. Sleep apnea is a medical condition and belongs with a clinician.

The Bottom Line

THC works on the part of sleep you can feel and degrades the part you can't. It gets you to sleep faster, it takes REM away, tolerance erodes the benefit within weeks, and quitting hands back a week or more of rebound that makes starting again look reasonable. That's a well-documented loop, and knowing the shape of it is more useful than any advice about what to do with the information. The cleanest reading of the evidence is that cannabis is a sedative with real costs and no demonstrated long-term sleep benefit, which is a different thing from it being useless to people who use it.

References

1. Feinberg I, Jones R, Walker JM, Cavness C, March J. Effects of high dosage delta-9-tetrahydrocannabinol on sleep patterns in man. Clinical Pharmacology and Therapeutics. 1975;17(4):458-466. PubMed: 164314
2. Gates PJ, Albertella L, Copeland J. The effects of cannabinoid administration on sleep: a systematic review of human studies. Sleep Medicine Reviews. 2014;18(6):477-487. PubMed: 24726015
3. Angarita GA, Emadi N, Hodges S, Morgan PT. Sleep abnormalities associated with alcohol, cannabis, cocaine, and opiate use: a comprehensive review. Addiction Science and Clinical Practice. 2016;11(1):9. PubMed: 27117064
4. Babson KA, Sottile J, Morabito D. Cannabis, Cannabinoids, and Sleep: a Review of the Literature. Current Psychiatry Reports. 2017;19(4):23. PubMed: 28349316
5. Vandrey R, Smith MT, McCann UD, Budney AJ, Curran EM. Sleep disturbance and the effects of extended-release zolpidem during cannabis withdrawal. Drug and Alcohol Dependence. 2011;117(1):38-44. PubMed: 21296508
6. Bolla KI, Lesage SR, Gamaldo CE, et al. Sleep disturbance in heavy marijuana users. Sleep. 2008;31(6):901-908. PubMed: 18548836
7. Bolla KI, Lesage SR, Gamaldo CE, et al. Polysomnogram changes in marijuana users who report sleep disturbances during prior abstinence. Sleep Medicine. 2010;11(9):882-889. PubMed: 20685163
8. Diep C, Tian C, Vachhani K, et al. Recent cannabis use and nightly sleep duration in adults: a population analysis of the NHANES from 2005 to 2018. Regional Anesthesia and Pain Medicine. 2022;47(2):100-104. PubMed: 34873024
9. Suraev AS, Marshall NS, Vandrey R, et al. Cannabinoid therapies in the management of sleep disorders: a systematic review of preclinical and clinical studies. Sleep Medicine Reviews. 2020;53:101339. PubMed: 32603954
10. Winiger EA, Ellingson JM, Morrison CL, et al. Sleep deficits and cannabis use behaviors: an analysis of shared genetics using linkage disequilibrium score regression and polygenic risk prediction. Sleep. 2021;44(3):zsaa188. PubMed: 32935850

These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.