Ask people why they use cannabis and sleep lands near the top of the list, right next to pain. It works, in the narrow sense that matters at 1am: you feel drowsy, you fall asleep faster, and the lights go out.
Then you wear something that measures sleep objectively for a few months, and the story gets more complicated. The nights you used look fine on total sleep time. The thing that quietly changes is the architecture underneath — and the mornings after you stop.
This post is about what the research actually shows for THC, CBD, sleep stages and heart rate variability. No moralising, no "just don't". Just what the data says, what your watch can and can't show you, and a two-week test you can run on yourself if you want a real answer instead of a guess.
TL;DR: THC reliably helps you fall asleep faster and can keep you asleep, but it suppresses REM sleep and increases REM latency — which is why heavy users report fewer dreams and then vivid, intense dreams when they stop. Subjective reports and objective sleep studies disagree: people rate their sleep as better while polysomnography often shows more wake time after sleep onset and lower sleep efficiency in chronic users. Tolerance builds within days to weeks, and withdrawal brings a well-documented rebound insomnia that makes quitting feel impossible — often the sleep problem people are treating is the withdrawal. On the cardiovascular side, THC raises heart rate acutely and cannabis smoking carries measurable cardiovascular risk, which is relevant because you're taking it right before the longest stretch of the day. CBD does not appear to suppress REM the way THC does, but the sleep evidence for CBD is much weaker than the marketing suggests, and low doses can be wake-promoting. Your watch will probably not show much on total sleep — look at REM percentage, overnight HRV, and resting heart rate instead.
YouTube: Related video
Why so much sleep trouble ends up in a joint or a vape
Cannabis is one of the most widely used substances on the planet, and sleep is one of the three most common reasons people give for using it — along with pain and anxiety. Dispensary surveys regularly put sleep in that top three, and it's the reason a lot of people who would describe themselves as "not really a cannabis person" end up with a gummy next to their bed.
The appeal is obvious and real. THC has sedative properties. It shortens sleep onset latency — the time it takes to fall asleep — and for someone who has spent two hours staring at the ceiling, that is not a small thing. It also tends to reduce the "wired but tired" feeling that keeps anxious people awake.
The confusion starts because "I fall asleep faster" and "I sleep better" are not the same claim. Falling asleep is one variable out of many, and it happens to be the one you subjectively feel most clearly.
What actually happens to your sleep stages
Sleep is not one state, and different drugs hit different parts of it.
REM sleep is the main casualty. The literature on cannabis and sleep is fairly consistent on this: THC reduces time spent in REM sleep and increases REM latency — the delay before your first REM period. A 2017 review by Babson and colleagues in Current Psychiatry Reports summarised decades of work and landed on REM suppression as one of the more reproducible findings, alongside mixed and often null results for slow-wave sleep.
That's worth sitting with, because REM sleep is where a lot of emotional processing and memory consolidation happens. REM suppression isn't automatically a catastrophe — plenty of drugs do it — but it does mean that "eight hours of sleep" on a THC night and eight hours without are not the same eight hours.
Slow-wave sleep (deep sleep) is murkier. Some trials show small increases, some show nothing, some show a reduction with chronic use. The effect depends heavily on dose, formulation, THC-to-CBD ratio, route of administration and how long the person has been using, which is exactly why the studies disagree with each other.
Tolerance develops fast. The sedative effect that made sleep easy in week one is noticeably smaller in week three. This is where the dose escalates, not because the person is chasing a high, but because they're chasing the same night of sleep they used to get at half the amount.
The subjective/objective gap (the part almost nobody knows)
Here's the finding that most surprises people. When you ask regular cannabis users about their sleep, they usually report it improved. When you put them in a lab and measure with polysomnography — electrodes, brain waves, the actual gold standard — the picture often looks worse: more wake time after sleep onset, lower sleep efficiency, and a higher proportion of light N1 sleep.
In other words: you feel like you slept well, and the recording shows a more fragmented night. This is a known phenomenon in sleep research and it shows up with alcohol too — the difference is that with weed, the subjective improvement is bigger and more convincing, because it fixes the falling-asleep part, which is the bit you experience consciously.
The same gap shows up in your dream recall. If you use most nights, you may not have noticed that you stopped remembering dreams. That's not a personal quirk. It's REM suppression doing its job.
The withdrawal trap — the most important section on this page
This is where cannabis and sleep get tangled in a way that's genuinely hard to untangle without data.
Stop using after a period of regular nightly use and you get a well-documented withdrawal syndrome: insomnia, vivid and often disturbing dreams, night sweats, irritability, restlessness, and reduced appetite. NIDA's research summary covers this directly, and it's one of the reasons cannabis use disorder is so sticky — sleep is the first thing to go when people try to stop.
Two things are going on at once:
- REM rebound. Your brain has been running a REM deficit. When THC clears, it tries to collect. That produces longer, more intense REM periods, which is why the dreams go from "I don't remember dreaming" to wildly vivid in the space of a couple of nights.
- Rebound insomnia. Sleep onset gets harder precisely because the thing that was helping you fall asleep is gone. This typically lasts one to two weeks, sometimes a bit longer with heavy, long-term use.
Which leads to the question worth asking honestly: when you lie awake on a night you don't use, is that your baseline insomnia, or is it withdrawal? For a lot of regular users it's mostly the second one. And that matters, because withdrawal insomnia resolves on its own in a couple of weeks with no treatment, while the belief that "I can't sleep without it" can become permanent.
What your wearable can actually show you
Your Apple Watch or ring cannot detect THC. It also won't flag a "bad" night on a night where you used, because total sleep time often looks similar-ish. So don't judge this by your sleep score.
The metrics that actually move:
| What to look at | What usually happens with evening use |
|---|---|
| Sleep onset latency | Shorter — often dramatically |
| REM percentage | Lower, sometimes well below your normal range |
| Wake after sleep onset | Equal or higher (the subjective report often disagrees) |
| Overnight HRV | Often suppressed — THC raises heart rate and shifts the balance away from parasympathetic tone |
| Resting heart rate | Typically elevated overnight and the next morning |
| Respiratory rate | Can shift slightly; watch for anything unusual |
| Dream recall (self-report) | Down, often to near zero |
The AHA's 2020 scientific statement on cannabis and cardiovascular health is worth knowing about here. It documents that cannabis — THC specifically — produces acute tachycardia (a rise in heart rate), can cause orthostatic hypotension, and that smoked cannabis carries risks similar in kind to tobacco smoke, plus some arrhythmia signals. None of that is a reason to panic over one edible, but it is a reason to care about whether you're using it every night before the longest stretch of stillness in your day.
Your watch isn't a diagnostic tool and it can't tell you why HRV dropped. What it can do is show you the pattern across weeks — and patterns are where honesty lives. If your overnight HRV is consistently 8-10 ms below your own baseline and your resting heart rate is up 4-5 bpm on weeks you use, that's a signal about the cost, whatever your subjective sleep rating says. For more on reading these signals without overreacting to single nights, see what it means when HRV is down and resting heart rate is up.
What about CBD?
CBD is the cannabinoid everyone sells for sleep, and the honest summary of the evidence is: promising, heavily under-studied, and not nearly as well supported as the marketing implies.
A few things the research does suggest:
- CBD does not appear to suppress REM sleep the way THC does. That's the single biggest argument in its favour, and it's based on a small number of studies, not a solid body of work.
- The dose-response looks like a U or J curve. Low doses (roughly under 20-25 mg) have been observed to be wake-promoting in some people. Sedation tends to show up at higher doses. So "take a small amount of CBD before bed" can genuinely backfire.
- A lot of the benefit may be indirect. CBD has reasonably good evidence for anxiety, and if your sleep problem is an anxious, ruminating brain, reducing that could plausibly improve sleep without acting on sleep architecture at all.
- Product labelling is a mess. Independent testing has repeatedly found CBD products containing more or less than labelled, and some containing THC when they claimed not to.
There's also an important legal and medical caveat: cannabis and CBD interactions with prescription medications — including common sleep and anti-anxiety drugs, blood thinners and antidepressants — are real, and this is a conversation to have with a pharmacist or doctor, not a blog post.
Two weeks, one variable: how to test this on yourself
If you want a real answer instead of vibes, here's a protocol simple enough to actually finish.
Week 1 — baseline. No cannabis. Same bedtime window (aim for a 30-60 minute band), same caffeine cutoff, no alcohol. Let the watch record normally. Note how you feel at 10am each day, not just at 7am.
Week 2 — your normal pattern. Same sleep window, same caffeine cutoff, still no alcohol. Use cannabis exactly the way you normally do — same product, same dose, same timing. Don't change anything else.
Then compare 7-day averages, not individual nights. Specifically:
- REM sleep percentage
- Wake after sleep onset (or "time awake" in your sleep app)
- Sleep onset latency
- Overnight HRV and morning resting heart rate
- Sleep score, for reference only
- Your 10am energy, scored honestly
Two rules for this to mean anything: keep the sleep window fixed so you're not comparing a 9-hour night to a 6-hour one, and don't change two things at once. If you're also cutting caffeine, you've learned nothing about the cannabis.
Also worth knowing: your wearable's REM detection isn't clinical grade. It's directionally useful and good for spotting trends, but if your REM percentage drops from 20% to 12% on use weeks consistently, that's a real signal even if the absolute number is a bit off. How sleep scores are calculated explains why the single score number is a worse guide than the components behind it.
If you want to cut back or stop
A few things that make it less brutal:
- Taper, don't stop cold. Rushing a taper makes the rebound insomnia worse and is the most common reason people give up. Reduce dose or frequency in steps over one to two weeks.
- Expect one to two weeks of rough sleep. Not "something is wrong with you". Rebound insomnia is the expected withdrawal path and it resolves.
- Protect the wake time above all else. During rebound, the temptation is to sleep in. Sleeping in pushes your circadian rhythm later and drags the whole thing out. Fixed wake time, morning light, no long naps.
- Have a fallback that isn't another sedative. For chronic insomnia, CBT-I is the first-line treatment with the strongest evidence base — better long-term results than sleep medication. It's not a soft option, it's the treatment that works. Our guide to CBT-I for insomnia covers what it involves.
- If you were using cannabis to mask a real problem — chronic pain, PTSD nightmares, anxiety — that's exactly the situation where a clinician should be involved, because there are treatment options with better evidence for those specific conditions.
And if alcohol is part of the evening routine too, note that it stacks in the same direction: it also suppresses REM, also raises overnight heart rate, and also produces rebound awakening. Two sedatives that both hurt sleep architecture don't cancel each other out. They compound. What alcohol does to your HRV the next morning is the direct comparison.
The short version
- THC shortens the time it takes to fall asleep. That's the part that works and it's real.
- It suppresses REM sleep and increases REM latency. Heavy users stop remembering dreams and often don't connect the two.
- Self-reported sleep improvement and objective sleep quality regularly disagree — more wake time, lower efficiency on paper.
- Tolerance builds in days to weeks, which is how the dose creeps up.
- Withdrawal brings rebound insomnia plus vivid dreams for one to two weeks. A lot of "I need it to sleep" is actually the withdrawal cycle, not baseline insomnia.
- CBD doesn't look like it suppresses REM the way THC does, but the evidence is thin, low doses can be wake-promoting, and label accuracy is poor.
- Your watch won't flag THC nights on total sleep. Watch REM percentage, overnight HRV and resting heart rate instead — and compare 7-day averages, never single nights.
The point of this piece isn't to tell you to stop. It's that "it helps me sleep" and "it improves my sleep" are different statements, and only one of them survives being measured. You're already wearing the measurement.
This isn't medical advice, and cannabis laws and medical guidance vary by where you live. If you're using it to manage a health condition or taking prescription medication, talk to a clinician.
Century AI helps you understand your body with a daily health score, recovery score, and sleep insights — using the watch you already wear.