Effects
Mixing Cannabis and Alcohol: Sleep, Dosing and Risks
A beer with dinner and a joint after it is an ordinary Friday for a lot of people. They do not cancel out. Alcohol pushes blood THC higher, both substances flatten REM sleep, and the second half of the night pays for the first. Here is what mixing them does to a night's sleep, how to time each one, and when the honest answer is to skip one.
The short version
- Alcohol and cannabis both get you under faster and both take REM sleep away. The cost lands in the back half of the night, when alcohol fragments sleep and the suppressed REM comes back all at once.
- Drinking alongside cannabis raises how much THC reaches your blood. In a controlled 2015 trial, a low dose of alcohol lifted median peak blood THC from 42.2 to 67.5 µg/L on the higher cannabis dose.
- Stop drinking 3 hours before bed. Your liver clears roughly 1 standard drink per hour and nothing speeds that up, so 3 drinks finished at 8pm are close to gone by an 11pm bedtime.
- Smoked or vaporized cannabis peaks around 10 minutes in and fades over 2 to 3 hours. An edible takes 30 to 120 minutes to arrive and is still working 6 to 8 hours later, which is the difference between falling asleep and waking up groggy.
- Start at 2.5mg THC if you are swallowing it, or 1 to 2 inhalations if you are smoking. A 1:1 product at 2.5mg THC and 2.5mg CBD is the gentler version of the same dose.
- Nightly use of either builds tolerance to the sleep effect within weeks. Stopping cannabis brings 1 to 2 weeks of broken sleep and unusually vivid dreams, which is REM rebounding rather than anything going wrong.
Alcohol raises how much THC reaches your blood
The interaction is not a vibe. In a controlled trial published in Clinical Chemistry in 2015, Hartman and colleagues gave 32 adult cannabis users either placebo or low-dose alcohol 10 minutes before vaporized cannabis, targeting a peak breath alcohol concentration of about 0.065%.
With placebo alcohol, median peak blood THC was 32.7 micrograms per liter (µg/L) on the low cannabis dose and 42.2 µg/L on the high dose. With alcohol on board, those peaks rose to 35.3 and 67.5 µg/L. Levels of 11-hydroxy-THC, the more potent metabolite, rose too.
The practical translation: the same joint delivers more when you have been drinking. If you know your dose from sober nights, that number is wrong on a night with 3 drinks in you, and it is wrong in the direction that catches people out.
Order matters less than people assume, but it is not nothing. Small studies have reported that cannabis taken first can slow alcohol absorption and lower peak breath alcohol readings, which is not the same as being less drunk. It just means the reading arrives late.
Then there is the greening out risk. Nausea, sweating, pallor, dizziness and vomiting are the common shape of a bad crossfade, and THC's anti-nausea effect can blunt vomiting at exactly the point where vomiting is the body's protection against alcohol poisoning.
Sleep runs in 4 stages, and the 2 drugs hit different ones
Sleep is not one block. It cycles through 3 non-REM stages and REM roughly every 90 to 110 minutes, 4 to 6 times a night, and the mix changes as the night goes on.
| Stage | Share of a normal night | What is happening | Where it sits |
|---|---|---|---|
| N1 | 2–5% | Drifting off, 1–7 minutes, easily woken | Start of each cycle |
| N2 | 45–55% | Body temperature drops, sleep spindles, heart rate slows | Throughout |
| N3 (deep, slow wave) | 13–23% | Hardest to wake from, growth hormone release, physical repair | Front-loaded into the first third |
| REM | 20–25% | Dreaming, brain activity near waking, muscles paralysed | Back-loaded into the last third |
The front-loading is the whole reason timing matters. N3 stacks into the first 2 cycles, while REM periods get longer as the night runs, so the final REM stretch before your alarm can last 30 to 60 minutes on its own.
That means a substance affecting only the first half of the night still leaves most REM intact. Anything that wrecks the back half takes almost all of it, and anything that suppresses REM directly takes it wherever it sits.
Numbers
A 7.5 hour night is 450 minutes. In stages, that comes out as:
- N3 deep sleep 60 to 105 minutes, and most of it inside the first 3 hours
- REM 90 to 112 minutes, split across 4 to 6 periods that lengthen as the night runs
- First REM period 5 to 10 minutes, arriving 70 to 110 minutes after lights out
- Last REM period 30 to 60 minutes, sitting in the 90 minutes before your alarm
- Cutting the night to 6 hours Costs roughly 35 REM minutes and almost no N3, because the part you remove is the end
Alcohol: asleep in 20 minutes, awake at 3am
Alcohol works as a sedative on the way in. It shortens the time it takes to fall asleep, and at moderate doses it increases N3 deep sleep across the first 2 cycles, which is why a nightcap genuinely feels like it works.
It suppresses REM in that same stretch and pushes the first REM period later. Then blood alcohol falls, the sedation lifts, and the second half of the night turns into the bill.
What that bill looks like: frequent brief awakenings, lighter sleep, an elevated heart rate, sympathetic nervous system activation, and REM arriving all at once in a rebound that often comes with intense or unpleasant dreams. Survey data puts a sleep complaint in front of nearly 90% of people who drink in the evening.
There are 2 mechanical problems on top of that. Alcohol suppresses vasopressin, so your kidneys produce more urine and a bathroom trip becomes likely. It also relaxes the muscles that hold the upper airway open, which makes snoring louder, apnea episodes longer and oxygen dips deeper.
Numbers: alcohol by the clock
1 standard drink 14g of pure alcohol: 12oz of beer at 5%, 5oz of wine at 12%, or 1.5oz of spirits at 40% Peak blood alcohol 30 to 90 minutes after the last sip Clearance About 1 standard drink per hour, and coffee, food and cold showers do not change it 3 drinks finished at 8pm Close to zero by 11pm The rebound window Roughly hours 4 to 8 of the night, which is when REM should be at its longest Nightcap tolerance The sleep-onset benefit fades within days to 2 weeks of nightly use
The standard drink figures come from the NIAAA definition, and they matter because a 16oz craft beer at 7% is closer to 2 drinks than 1. Most people who think they had 3 had 4 or 5.
Cannabis: faster to sleep, less dreaming, a tolerance clock
Sleep is one of the most common reasons people buy cannabis. Roughly 48% of medical cannabis patients report using it for insomnia, and in surveys of adult users around 77% of men and 64% of women say they have tried it for sleep at some point.
Acutely, THC shortens sleep latency for most people, typically pulling 15 to 30 minutes off the time it takes to drop off. It suppresses REM in a dose-dependent way, so bigger doses take more dreaming.
The effect on deep sleep is less settled. Some studies show a modest increase in N3 at low doses, others show a decrease at higher doses, and the effect is not consistent enough to sell as a benefit.
Here is the part that gets left out. Most of the sleep research on cannabis uses small samples, runs for days or weeks rather than months, and often studies people without a diagnosed sleep disorder. Short-term benefit is well documented. A durable long-term fix is not.
Tolerance is the reason. Nightly use erodes the sleep effect within 1 to 4 weeks for most people, and the standard response is to raise the dose, which works for a while and then does not.
Cannabis or alcohol: which is worse for sleep
This is the question underneath most searches on the topic, so here is the direct comparison, including the night you have both.
| Alcohol (2–3 drinks) | THC (5–10mg, or 2–3 inhalations) | Both on the same night | |
|---|---|---|---|
| Time to fall asleep | Shorter, often under 20 minutes | Shorter, typically by 15–30 minutes | Shortest, and the least controllable |
| First half of night | More N3 deep sleep | Modest N3 change, evidence mixed | Heavily sedated, low arousal |
| REM | Suppressed early, rebounds later | Suppressed, dose-dependent | Suppressed by both, bigger rebound |
| Second half of night | Fragmented, repeated brief awakenings | Usually intact if the dose was low | Fragmented, with THC still active |
| Breathing | Worse: airway muscles relax, apneas lengthen | Neutral to unclear | Worse, and alcohol is the reason |
| Bathroom trips | More likely, diuretic effect | No effect | More likely |
| Next morning | Headache, dehydration, flat mood | Grogginess, mostly after edibles | Both, and heavier than either alone |
| Tolerance to the sleep effect | Days to 2 weeks of nightly use | 1–4 weeks of nightly use | Faster, since 2 systems are being loaded |
| Dependence risk | High and well documented | Real, and concentrated in daily users | Higher than either alone |
| Acute overdose danger | Yes, alcohol poisoning is fatal | No reported fatal THC overdose | Alcohol remains the danger |
For a single night of sleep architecture, a low THC dose does less damage than 3 drinks, mostly because it leaves the back half of the night alone. That is a narrow claim about one night, not a verdict on the 2 substances.
Neither is a treatment for insomnia. Both trade the first half of the night against the second, and both stop working at the original dose if you use them every night.
The night you have both: what actually happens
You have 2 sedatives arriving together, and one of them is raising the blood level of the other. You fall asleep faster than on either alone, and you fall asleep having lost track of how much of each is still to arrive.
The first 2 cycles are heavy. Then alcohol clears somewhere around hour 4, the rebound arousal hits, and unlike a straight drinking night the THC has not finished, so the fragmented back half also has REM suppression sitting on top of it.
Direct polysomnography studies of people using both are thin on the ground, so this picture is assembled from what each does on its own plus the pharmacokinetic interaction. What is not in doubt is that the impairment stacks: combined use produces more impairment on driving and coordination tasks than either substance alone at the same doses.
Worked example: 3 drinks and a 5mg gummy
7:00pm First drink with dinner 9:00pm Third drink finished. Blood alcohol still climbing. 9:30pm 5mg gummy. Alcohol on board, so more of the THC will reach the blood than on a sober night. 10:30pm Both arriving together. Sedation heavier than expected, judgement about taking more is poor. 11:15pm Asleep in under 10 minutes, deeper than usual for the first 2 cycles. 3:00am Alcohol at zero. Rebound arousal, heart rate up, bathroom trip likely. 3:30am to 6:30am Light, broken sleep. THC still suppressing REM. 7:00am Alarm. 7 hours 45 minutes in bed, maybe 5 of it useful.
The same evening with the last drink at 8pm and no gummy costs about 20 extra minutes falling asleep and returns most of the back half.
Timing: how many hours before bed for each
Timing is the single highest-leverage change available, and it costs nothing.
| Substance and format | Take it this far before bed | Why |
|---|---|---|
| Alcohol | 3 hours, longer past 3 drinks | Blood alcohol near zero at lights out avoids the rebound landing mid-night |
| Smoked flower | 15–30 minutes | Peaks around 10 minutes, tapers over 2–3 hours |
| Vaporized flower or cart | 15–30 minutes | Same curve, less throat irritation |
| Tincture held under the tongue | 30–45 minutes | Part absorbs sublingually, part behaves like an edible |
| Edible or capsule | 60–120 minutes | Onset is wide and the tail runs 6–8 hours |
If you are having both on the same night, put the 3 hours between the last drink and bed, and choose the inhaled route over an edible. That way the alcohol is gone before the rebound window and the THC is largely finished before the last REM stretch.
The common mistake is the opposite: a drink at 10:30pm and an edible at 11pm, which puts the alcohol rebound and the edible peak in the same 3 hours.
Delivery method changes the drug
The same milligram number behaves like a different drug depending on how it enters you. That is not marketing, it is absorption.
| Format | Onset | Peak | Duration | Bioavailability | What it suits |
|---|---|---|---|---|---|
| Smoked flower | 1–5 min | ~10 min | 2–3 hours | Up to 56% | Falling asleep |
| Vaporized | 1–5 min | ~10 min | 2–4 hours | Up to 56% | Falling asleep, gentler on the throat |
| Tincture, sublingual | 15–45 min | 60–90 min | 4–6 hours | Between the 2 | Measurable dosing, middle-length effect |
| Edible, swallowed | 30–120 min | 2–3 hours | 6–8 hours | 4–20% | Staying asleep, at the cost of morning fog |
| Capsule | 45–120 min | 2–3 hours | 6–8 hours | 4–20% | Same as an edible, no sugar |
Swallowed THC passes through the liver before it reaches your brain, and a large share of it converts to 11-hydroxy-THC, which is more potent than THC itself. Less gets absorbed, and what does arrive hits harder and lasts longer.
That is why 5mg eaten is not comparable to 5mg inhaled, and why an edible taken at 11pm can still be working at 5am. For staying asleep that is the point. For a 7am alarm it is the problem.
THC, CBD and CBN: the ratio is the lever
THC does the sedating and THC does the REM suppression. It is also what produces anxiety and a racing heart above your personal threshold, which is why the dose that helps and the dose that ruins the night can be 5mg apart.
CBD is not the sedative most people assume. Low doses can be mildly alerting, and sedation only shows up at doses far above the 10mg or 25mg in a typical gummy. Its job in a sleep product is taking the edge off THC's anxiety, not adding sleepiness of its own.
The ratio on the label is what you are actually buying:
- 1:1 Equal THC and CBD, often 5mg of each. The standard step for someone who finds straight THC too much.
- 2:1 or 4:1 CBD-forward. Milder, and reliably weaker as a sedative.
- 20:1 CBD-dominant, often 20mg CBD to 1mg THC. Barely intoxicating and unlikely to knock anyone out.
- THC only Strongest sedation, strongest REM suppression, highest chance of an anxious night at the wrong dose.
CBN is the cannabinoid sleep products are marketed on, and the evidence behind it is weak. The small human studies from the 1970s found CBN on its own was not meaningfully sedating and added little to THC. The sleep category on a Detroit dispensary menu is usually 3 or 4 CBN gummies, and the panel on them shows 2mg to 5mg of CBN sitting beside 3mg to 5mg of THC. The THC is doing most of the work.
What to buy and what to say at the counter
The number that matters on the label is milligrams per piece, not milligrams per package. A 100mg bag is 10 servings, and the 100mg is on the front for compliance rather than for you.
Colorado sets a single retail edible serving at 10mg of THC with 100mg per package, which is why most markets sell gummies 10 to a bag. Massachusetts shops work to a 5mg serving cap instead, so the same brand arrives in a smaller piece depending on which state line you are standing on.
Ratio products are a stock item in mature markets and a special order in newer ones. A Seattle dispensary will usually have a 1:1 tincture, a 4:1 gummy and a 20:1 capsule in the case at the same time, while a store in a younger market may carry one of the 3.
Numbers: what a sleep shelf costs
10-piece gummy pack, 10mg THC each $15 to $30 before tax CBN sleep gummies, 5mg THC plus 2 to 5mg CBN per piece $20 to $35 1:1 tincture, 30ml, 300mg THC and 300mg CBD $35 to $60 20:1 CBD capsules, 30 count $40 to $70 Eighth of flower, 3.5g $25 to $50 depending on the market Tax at the register 10% to 37% on top of the shelf price, set by the state
At the counter
Ask these 4 things and you will leave with the right product rather than the one on the end cap:
- "What is the lowest milligram per piece you carry, and is it scored?" A 5mg piece you can halve is a 2.5mg starting dose. A moulded 10mg chocolate is not divisible.
- "Do you have a 1:1 in this format?" If straight THC makes you anxious, this is the single most useful substitution on the menu.
- "What is the CBN number on your sleep gummies, and what is the THC number next to it?" If the THC is 5mg and the CBN is 2mg, you are buying a 5mg THC gummy with a sleep label.
- "Anything that arrives fast and leaves fast?" That question gets you pointed at flower, a vape or a sublingual tincture instead of a 6-hour edible.
Tolerance, quitting, and the vivid dreams
Nightly THC stops working for sleep. Most people notice within 1 to 4 weeks that the dose that used to work does not, and the standard move is to double it, which buys another few weeks.
Stopping is where the cost becomes visible. Cannabis withdrawal is real and sleep sits at the centre of it: difficulty falling asleep, more awakenings, and strange, unusually vivid dreams.
The timeline is fairly consistent. Symptoms start 1 to 3 days after the last use, peak somewhere around days 2 to 6, and mostly resolve inside 1 to 2 weeks, though sleep can take 3 to 4 weeks to settle.
The dreams are the clearest signal of what the drug was doing. REM that was suppressed for months comes back at once, and the brain runs the backlog. It is uncomfortable and it is temporary.
Alcohol follows the same shape. Rebound insomnia after stopping a nightly drink habit is common and it also settles within a couple of weeks, which is exactly long enough for people to conclude they cannot sleep without it.
At the counter
If you are trying to step down rather than stop, buy the smaller piece rather than trying to eat less of the big one. A pack of 2.5mg gummies costs the same as a pack of 10mg gummies and makes a taper possible.
Say: "I am cutting back and I want the lowest dose you stock, not the best value per milligram." Budtenders hear the value question all day and almost never hear this one.
Where the stakes are highest: pain, PTSD, apnea, restless legs, anxiety
The evidence is not evenly spread. Some conditions have real support behind cannabis, some have almost none, and one has an active recommendation against it.
- Chronic pain carries the strongest evidence base, and the sleep improvement usually follows the pain relief rather than any sedative effect. This is the group most likely to get a genuine result.
- PTSD and nightmares have small trials behind them, mostly using synthetic THC, showing reduced nightmare frequency. Alcohol is the worst possible pairing here, because the REM rebound window is exactly when nightmares run.
- Sleep apnea is the clear negative. The American Academy of Sleep Medicine's 2018 position statement advises against medical cannabis for obstructive sleep apnea, citing unreliable delivery and insufficient evidence. Alcohol independently makes apnea worse.
- Restless legs syndrome has case reports and nothing stronger. Alcohol reliably worsens it, which is the more actionable half of that sentence.
- Insomnia gets short-term benefit and no evidence of a lasting fix. Cognitive behavioural therapy for insomnia is the treatment with the durable track record.
- Anxiety cuts both ways. Low THC doses calm most people and higher doses do the opposite in the same person, and alcohol adds its own rebound anxiety in the small hours.
Who should not mix them, and when to see a doctor
Some combinations are not a matter of dose. These are the cases where the answer is no rather than less.
- Anyone on sleeping pills, benzodiazepines or opioids. Zolpidem, eszopiclone, benzodiazepines and opioids are central nervous system depressants, and alcohol or THC on top is additive. Do not add alcohol at all, and do not add cannabis without asking the prescriber.
- Untreated or suspected sleep apnea. Alcohol lengthens apnea events and deepens oxygen dips, and cannabis is not the alternative.
- Anyone driving within the next several hours. Combined impairment exceeds either substance alone, and an edible can leave you unfit to drive 6 to 8 hours later.
- Pregnancy and breastfeeding. Neither substance has a safe established level.
- A personal or family history of psychosis. THC is the relevant risk here, and higher doses are the relevant version of it.
- Alcohol use disorder, or drinking to fall asleep. Adding a second sedative to a nightly drink habit builds 2 dependencies where there was 1.
The threshold for a doctor is concrete rather than vague. Trouble falling or staying asleep at least 3 nights a week for 3 months meets the clinical definition of chronic insomnia and is worth a proper assessment rather than a shelf product.
Book the appointment sooner if a bed partner reports that you stop breathing or gasp during the night, if you cannot fall asleep at all without a drink or a dose, or if you have raised your dose twice in 3 months to get the same effect.
Common questions
What happens if you mix weed and alcohol?
Alcohol increases how much THC reaches your blood, so the same cannabis dose hits harder than it does sober. The 2015 Hartman trial recorded median peak blood THC rising from 42.2 to 67.5 µg/L on the higher dose when low-dose alcohol was taken first. Practically that means stronger effects than expected, worse coordination than either alone, and a higher chance of nausea and vomiting.
Is weed or alcohol worse for sleep?
For a single night, 2 or 3 drinks does more damage than a low THC dose, because alcohol fragments the entire back half of the night while a small dose of THC mostly leaves it alone. Both suppress REM and both stop working at the original dose within weeks of nightly use. Neither is a treatment for insomnia.
How many hours before bed should you stop drinking?
3 hours, and longer if you had more than 3 drinks. Your liver clears about 1 standard drink per hour, and the goal is a blood alcohol close to zero when you get into bed, so the rebound arousal does not land at 3am. Nothing accelerates clearance, so the only variable you control is when you stop.
Does weed affect REM sleep?
Yes, THC suppresses REM sleep in a dose-dependent way, so larger doses take more of it. REM is normally 20% to 25% of the night and is concentrated in the last third, which is why an edible still working at 5am costs more dreaming than a joint at 11pm. CBD does not suppress REM the way THC does.
Why do I get vivid dreams when I stop smoking weed?
Because the REM sleep that THC was suppressing returns all at once. It is called REM rebound, it usually starts within 1 to 3 days of the last use, peaks around days 2 to 6, and settles inside 1 to 2 weeks. The dreams are intense and sometimes unpleasant, and they are a normal part of the sleep architecture recovering.
Is a weed hangover real?
Yes, though it looks different from an alcohol hangover. The usual pattern after an edible is grogginess, mild brain fog and a slow start the next morning, driven by how long oral THC stays active rather than by dehydration. It is far more common with 10mg or more taken late at night than with inhaled cannabis, which has mostly cleared by bedtime plus 3 hours.
Is cannabis safer than alcohol?
On acute overdose, yes: there are no reported deaths from THC alone, while alcohol poisoning kills. On everything else the comparison is more mixed, since cannabis carries dependence risk, cognitive effects and a psychosis risk in vulnerable people, and long-term studies on modern high-THC products are limited. Safer on one axis is not safe.