Many people who use marijuana for sleep are not actually struggling to fall asleep. They can drift off within 15 or 20 minutes, only to wake repeatedly during the night, remain awake for long stretches at 2 or 3 a.m., or wake several hours earlier than intended. That pattern is often described as sleep-maintenance insomnia, and it is clinically different from sleep-onset insomnia—the difficulty initiating sleep in the first place. The distinction matters because a substance that makes someone feel sleepy at bedtime does not necessarily keep that person asleep throughout the night.
Cannabis has become a popular self-directed sleep aid, and there is legitimate scientific interest in THC, CBD, CBN, and combinations of cannabinoids for insomnia. Some controlled studies have reported improvements in perceived sleep quality, total sleep time, or the time required to fall asleep. Evidence specifically addressing nighttime awakenings, however, is much more inconsistent. The 2025 VA/Department of Defense clinical practice guideline for chronic insomnia reviewed the available cannabinoid evidence and ultimately suggested against cannabis or cannabis derivatives for treating chronic insomnia, concluding that the quality of evidence remained very low.
Sleep Maintenance Insomnia vs Sleep-Onset Insomnia
The National Library of Medicine describes insomnia as trouble falling asleep, staying asleep through the night, or waking too early. Those problems can occur together, but many patients predominantly experience one pattern. Sleep-onset latency, usually abbreviated SOL in research papers, measures approximately how long it takes someone to fall asleep. Sleep maintenance is commonly evaluated using wake after sleep onset, or WASO—the cumulative amount of time a person spends awake after initially falling asleep. Researchers also measure total sleep time and sleep efficiency, the proportion of time spent in bed that is actually spent sleeping.
This distinction is particularly important when evaluating marijuana effects. THC can produce noticeable relaxation, sedation, or altered perception shortly after use, so a person may reasonably conclude that it “helps insomnia” because bedtime feels easier. But feeling sedated before bed and maintaining consolidated sleep are not equivalent biological outcomes. A treatment could shorten sleep-onset latency by ten minutes while doing nothing to prevent several lengthy awakenings later in the night. For someone whose main problem is waking repeatedly rather than falling asleep, WASO and sleep efficiency are therefore more relevant endpoints than whether cannabis produces immediate sleepiness.
Can Marijuana Help You Stay Asleep?
One of the better-known randomized trials of cannabinoid treatment for insomnia tested a sublingual product called ZTL-101 containing THC, CBD, and CBN. Twenty-three participants completed the crossover trial. Compared with placebo, the cannabinoid formulation improved overall insomnia scores and self-reported sleep quality. Participants estimated that they fell asleep about eight minutes faster, while actigraphy showed approximately 10 fewer minutes of wake after sleep onset and about 33 additional minutes of total sleep time. Sleep efficiency also improved modestly.
Those findings are encouraging for sleep-maintenance insomnia because WASO actually improved rather than merely sleep onset. They remain preliminary, however. The trial was extremely small, lasted only two weeks per treatment period, and studied a specific standardized cannabinoid formulation rather than ordinary dispensary flower, gummies, vape cartridges, or oils. Thirty-six of the 40 mild adverse events recorded during treatment occurred while participants were receiving the cannabinoid product. A small short-term trial can identify a possible therapeutic signal, but it cannot establish whether nightly marijuana keeps people asleep safely and effectively for months or years.
Other research has produced almost the opposite result. In a study of 177 adults who underwent overnight sleep testing, cannabis use within three hours of bedtime was associated with greater wake after sleep onset. Median WASO was approximately 60.5 minutes among people using cannabis close to bedtime compared with 45.8 minutes in the comparison group. Frequent users also showed more nighttime wakefulness and lower sleep efficiency. Because this was observational research, it cannot prove that cannabis caused fragmented sleep; people with worse sleep may be more likely to use cannabis at night. Nevertheless, the findings challenge the common assumption that bedtime cannabis necessarily produces more continuous sleep.
Marijuana May Affect Falling Asleep Differently
The evidence for sleep initiation is somewhat more suggestive. A 2026 randomized placebo-controlled crossover trial from the Woolcock Institute of Medical Research in Sydney tested cannabinol, or CBN, in 20 adults with diagnosed insomnia. Researchers evaluated 30-mg and 300-mg oral doses. The primary outcome was wake after sleep onset. Neither dose significantly improved WASO compared with placebo. Yet the 300-mg dose did significantly shorten sleep-onset latency, improve subjective sleep quality, and reduce EEG arousals. In other words, the cannabinoid appeared to help participants fall asleep without clearly helping them stay asleep.
A 2026 meta-analysis of ten randomized trials involving more than 2,100 participants similarly found that cannabinoids produced an average reduction in sleep-onset latency of roughly 12 minutes while also improving several broader sleep measures. The same analysis reported improved total sleep time and sleep efficiency, but adverse effects—most commonly dizziness and dry mouth—were more frequent with cannabinoids. The authors concluded that cannabinoids show potential but emphasized the need for larger and longer-term trials.
Those results illustrate why asking simply whether “weed helps sleep” can obscure the real question. A person who lies awake for an hour before falling asleep could potentially value a modest reduction in sleep latency. Someone who falls asleep instantly but wakes four times every night needs evidence that a treatment reduces WASO and nighttime awakenings. At present, the research supporting cannabinoids for that second problem is notably less consistent.
THC, CBD and CBN Are Not the Same Sleep Aid
Marijuana’s sleep effects also depend on which cannabinoid is being considered. THC is intoxicating and can cause relaxation or drowsiness at some doses, although responses vary considerably. CBD does not produce the same typical high and is frequently marketed for sleep despite weak evidence when used by itself. A randomized 2024 trial from Swinburne University of Technology gave people with moderate-to-severe insomnia 150 mg CBD nightly for two weeks. CBD did not significantly improve insomnia severity, sleep-onset latency, sleep efficiency, or wake after sleep onset compared with placebo.
This is consistent with a 2025 systematic review and meta-analysis finding that CBD-only interventions did not significantly improve subjective sleep quality, whereas non-CBD cannabinoid formulations showed a stronger signal. Even those results need caution because cannabinoid trials vary enormously in product, THC concentration, CBD dose, patient population, study duration, and method of measuring sleep.
CBN has received increasing attention because it is commonly marketed as a “sleep cannabinoid.” The 2026 randomized trial gives that claim some scientific support regarding falling asleep, but not staying asleep: even a 300-mg dose failed to produce a significant reduction in WASO. That makes CBN an especially useful example of why cannabinoid marketing claims and clinical sleep-maintenance outcomes should not be treated as interchangeable.
Cannabis Can Change Sleep Architecture Without Improving Sleep Maintenance
Another complication is that sleep is not simply a state of unconsciousness. During a normal night, the brain cycles repeatedly through light non-REM sleep, deeper slow-wave sleep, and REM sleep. A drug can make someone feel sedated while also altering the normal structure of those cycles. A 2025 randomized pilot trial tested a single oral dose containing 10 mg THC plus 200 mg CBD in 20 patients with diagnosed insomnia using high-density EEG. The combination did not significantly improve wake after sleep onset. Instead, total sleep time actually decreased by about 25 minutes. REM sleep fell by nearly 34 minutes and REM latency increased by more than an hour.
A broader 2025 systematic review and meta-analysis of polysomnography research concluded that cannabis does not consistently improve sleep duration, latency, wake time, sleep efficiency, or sleep stages. Earlier studies suggesting strong suppression of REM sleep frequently involved small samples and unusually high THC doses. The modern evidence therefore does not support a simple model in which cannabis reliably deepens or stabilizes sleep.
This helps explain a common discrepancy in cannabis research: patients can report that they “slept better” while objective monitors find little improvement—or even more fragmented sleep. Subjective sleep matters; feeling rested is an important clinical outcome. But someone specifically trying to treat repeated nocturnal awakenings should recognize that sedation, perceived sleep quality, total sleep time, and objectively measured WASO can move in different directions.
Nightly Cannabis Use May Become Less Reliable Over Time
Short experimental trials also cannot fully answer what happens when marijuana becomes a nightly sleep aid. Tolerance can develop to some cannabis effects with repeated exposure, which may encourage increasing THC doses. Observational sleep-clinic research published in 2025 examined more than 1,400 patients and found that long-term daily cannabis use was associated with higher wake after sleep onset and poorer sleep efficiency after adjustment for numerous demographic, medical, medication, and sleep-related variables. Daily users had about 21 percent higher WASO than never-users in the adjusted analysis.
Again, observational research cannot prove that cannabis produced the sleep disruption. Insomnia itself may drive chronic cannabis use. But these results are important because they point in a different direction from the short-term sedation many users experience. Someone may find marijuana extremely effective for falling asleep tonight while gradually becoming dependent on it as part of the bedtime routine without necessarily achieving better long-term sleep maintenance.
Researchers at the University of Wyoming reported in 2026 that many people continue using cannabis for sleep despite insufficient evidence supporting it as a sleep aid. This pattern makes long-term studies particularly important because short-term trials cannot adequately measure tolerance, escalation, dependence, or how sleep changes after years of nightly exposure.
Cannabis Withdrawal Can Make Insomnia Temporarily Worse
People who use cannabis nightly for sleep can also discover that sleep deteriorates when they stop. The National Institute on Drug Abuse lists insomnia and disturbing dreams among recognized cannabis withdrawal symptoms following heavy or long-term use. That creates the potential for a self-reinforcing cycle: marijuana helps a person feel sleepy, regular use develops, stopping produces several nights of poor sleep, and the person interprets the withdrawal insomnia as evidence that marijuana is medically necessary for sleep.
Objective sleep research supports this phenomenon. Johns Hopkins-affiliated researchers studied daily cannabis users during short-term abstinence and found decreased total sleep time and sleep efficiency, along with increased sleep latency and changes in REM sleep compared with periods of cannabis use. Another study of heavy marijuana users found poorer total sleep time, slow-wave sleep, and sleep efficiency during early abstinence.
Withdrawal does not mean everyone using marijuana for sleep will become dependent. It does mean that a person who has used THC every night for months should not necessarily judge their natural baseline sleep from the first few nights after stopping. NIDA notes that withdrawal can occur after stopping or substantially reducing heavy or long-term cannabis use even in people who do not meet criteria for cannabis use disorder.
What Current Insomnia Guidelines Recommend
Mainstream sleep guidelines currently do not place marijuana alongside established insomnia treatments. The 2025 VA/DoD Clinical Practice Guideline for Chronic Insomnia Disorder and Obstructive Sleep Apnea specifically suggests against cannabis and cannabis derivatives for chronic insomnia. Its review included the recent CBD trial and concluded that the evidence remained very low quality, with no reliable improvement in objective wake after sleep onset, sleep efficiency, sleep-onset latency, total sleep time, or number of awakenings in the CBD study.
The American Academy of Sleep Medicine instead gives a strong recommendation to cognitive behavioral therapy for insomnia, or CBT-I. Importantly for sleep-maintenance insomnia, the evidence supporting CBT-I includes improvements not merely in falling asleep but also in wake after sleep onset. CBT-I commonly combines sleep restriction or sleep-compression strategies, stimulus control, cognitive therapy, and other methods designed to rebuild consolidated sleep rather than simply produce sedation at bedtime.
Medication options can also differ depending on the insomnia pattern. For example, the AASM pharmacologic guideline identifies certain treatments specifically for sleep-maintenance insomnia rather than assuming every hypnotic medication addresses both sleep onset and nighttime awakenings equally. That same principle should be applied when considering cannabis: whether something makes a person sleepy is not enough evidence that it treats the type of insomnia they actually have.
Final Thoughts on Marijuana for Sleep Maintenance Insomnia
Marijuana may help some people with insomnia, but the available research does not show that cannabis reliably prevents nighttime awakenings. A small trial of a THC/CBD/CBN formulation reduced wake after sleep onset by approximately ten minutes, and recent meta-analyses suggest cannabinoids may improve some overall measures of insomnia. At the same time, CBD alone has failed to improve WASO in a controlled trial, a 2026 CBN trial improved falling asleep without significantly improving staying asleep, and several objective sleep studies have associated cannabis exposure with more nighttime wakefulness or reduced sleep efficiency.
The distinction between falling asleep and staying asleep is therefore central. THC’s sedating effects may make the beginning of the night easier without guaranteeing consolidated sleep several hours later. Regular use can further complicate the picture through tolerance, dependence, withdrawal-related insomnia, and changes in sleep architecture. The National Center for Complementary and Integrative Health consequently describes the effects of cannabis and cannabinoids on primary sleep disorders as uncertain, even though sleep sometimes improves when cannabinoids relieve another problem such as chronic pain.
For someone who repeatedly wakes during the night, the most useful first question is not simply “Will marijuana make me sleepy?” It is why the awakenings are occurring. Sleep-maintenance insomnia can be associated with chronic insomnia disorder, obstructive sleep apnea, restless legs syndrome, pain, menopause, depression, alcohol use, medications, nocturia, circadian problems, and other conditions requiring different treatment. Medical cannabis research is evolving quickly, and particular cannabinoid formulations may eventually earn a defined role in insomnia treatment. For now, however, the strongest clinical evidence remains behind identifying the cause of fragmented sleep and using treatments such as CBT-I that have demonstrated durable improvements in both sleep initiation and sleep maintenance.
