Strange, vivid dreams after quitting weed are real and well documented: across 11 studies, an average of 34.4% of people reported them, with individual studies ranging from 5.1% to 78.0%. The popular explanation, that THC switches off REM sleep while you use, is far less settled than most articles admit.[1][4]

Key Takeaways

  • The dreams on stopping are the best-evidenced part. Reported rates average 34.4% across 11 studies, from 5.1% in a general survey to 78.0% among dependent daily users.[4]
  • REM shifts during abstinence were measured in a sleep lab. In 20 daily users, REM arrived sooner and took up a larger share of the night during three days without cannabis.[3]
  • "THC suppresses REM" is not established. Of six studies that measured sleep with EEG, one found REM increased, one found it decreased and four found no effect.[1]
  • The timing varies by study. One study saw strange dreams start around day 2 and peak around day 9; another put the peak for nightmares at about 1.5 days.[4]
  • No study here measured a dream. Dreams appear only as a self-rated checklist item, so "why they feel so intense" is inference, not measurement.[3][4]

Does weed stop you dreaming?

Nobody has shown that it does. If you rarely remember dreams while using regularly, the sleep research cannot yet tell you why: across six studies that measured sleep objectively, four found no change in REM sleep at all.[1]

It is worth separating two things that get blurred together. Dreaming is something the brain does during sleep. Dream recall is whether you remember any of it when you wake up. You can dream all night and recall nothing.

None of the studies behind this article measured dream recall in people who were using cannabis. What they measured was sleep staging: how long people spent in each stage of sleep, recorded with electrodes. That is a different question, and its answer is covered in the REM section below.

So if your dreams went quiet while you were a regular user, that is your experience. What has not been established is the mechanism behind it.

Why do you get vivid dreams when you stop smoking weed?

Strange and vivid dreams are one of the best-documented features of stopping cannabis. Pooled across 11 studies, 34.4% of people reported them (range 5.1% to 78.0%). In a sleep lab, REM sleep arrived sooner and took up more of the night during abstinence.[3][4]

The clearest measured evidence comes from a 2011 inpatient study. Researchers admitted 20 daily cannabis users (people using on at least 25 days a month, average age 29, 85% male) and recorded their sleep with polysomnography through three days of abstinence.[3]

Three changes were statistically significant. REM latency (the time from falling asleep to the first REM period) went down. The percentage of the night spent in REM went up. And ratings of "strange/wild dreams" on a withdrawal checklist went up. In the authors' words, REM share "increased and REM latency decreased during withdrawal in this study."[3]

Two limits matter. The study had no comparison group: the authors note that "a control group of non-cannabis users was not included." And it covered three days, which is not the whole period people describe dream changes over.[3]

How long do weed dreams last after you quit?

In the one study that tracked dreams day by day, strange dreams began around day 2 and peaked around day 9. Another study put the peak for nightmares at about 1.5 days. Reports of strange dreams lasting up to seven weeks have been observed in some people.[3][4]

Those two peaks (day 9 and day 1.5) do not agree, and averaging them would produce a number neither study found. The studies used different checklists, different wording ("strange dreams" in one, "nightmares" in the other) and different groups of people, which is a plausible reason for the gap.[4]

Notice too that dreams and general sleep trouble did not move together. In the same 2003 study, sleep difficulty started on day 1, peaked on day 2 and was back to baseline by day 12, while strange dreams were still building.[4]

If you are taking a break to bring your tolerance down, see our guide to why your usual amount stops working. If sleep disruption carries on or starts affecting your days, a clinician is a better next step than more reading.

StudyWhat was measuredFinding as reported
Vandrey 201120 daily users, inpatient, 3 days abstinent, sleep lab recordingREM latency decreased; REM percentage increased; "strange/wild dreams" increased
Jones 1976Inpatient withdrawalREM and eye movements increased during withdrawal; "wild dreams" went from 0% to 19%
Bolla 2008Sleep lab recording during abstinenceSleep efficiency and REM latency decreased; no effect on REM time
Bolla 2010Sleep lab recording during abstinenceTotal sleep time, REM time and sleep efficiency decreased throughout
Cohen-Zion 2009Sleep lab recording, adolescent usersNo effect on total sleep time, REM, stages 1 or 2, wake time, efficiency or sleep latency
Budney 2003Day-by-day symptom trackingStrange dreams: onset day 2, peak day 9. Sleep difficulty: onset day 1, peak day 2, baseline by day 12
Copersino 2006Symptom timingTrouble sleeping began at 2.74 days (SD 3.77) and lasted 43 days (SD 55)

Sources: Vandrey 2011[3]; all other rows as reported in Gates et al. 2016.[4] The large standard deviations in the Copersino row are a reminder of how much individual timelines varied.

Does THC actually suppress REM sleep?

Less clearly than almost every article on this subject claims. Reviewing the studies that measured sleep objectively, Gates and colleagues found REM increased in one, decreased in one and showed no effect in four. A 2025 meta-analysis found no consistent effect on sleep staging.[1][2]

The 2014 review read 39 publications. Only six measured sleep with EEG while people were given cannabis or THC, and the review's own summary is blunt: "REM sleep was reported to increase in one study, decrease in a second study, while four studies showed no effect." The authors concluded the mixed findings "preclude any definitive conclusion."[1]

The 2025 meta-analysis identified 18 studies and pooled nine. It found no consistent effect of cannabis administration on sleep staging, and noted that the early studies, run between 1972 and 1982, had between 2 and 11 participants each.[2]

Be careful not to overcorrect. Four studies finding no effect, most of them small, is weak and mixed evidence. It is not proof that THC leaves REM untouched. The honest summary is that the effect is not established.

StudyPeopleDose and routeNightsREM finding as reported
Barratt 197412 (8 dosed)0.2 mg/kg THC per joint, twice a day10No effect on stage 1, REM or time asleep
Hosko 19737Oral THC, two dose levels2, plus 5 placebo eachREM briefly increased then decreased at the high dose only; no net drug effect
Karacan 197664Usual use, average 9.2 joints a day8Sleep latency to stage 1 and REM time increased
Pranikoff 197320Dose unclear2No effect on time asleep, stage 1, stage 3 or REM time
Tassinari 199990.7 to 1.4 mg/kg oral THC, single dose1Stage 2 increased; deep sleep and REM decreased
Nicholson 2004815 mg THC1No effect on time asleep, stages 1 to 4, awakenings or efficiency

Source: Gates et al. 2014, Table 1.[1] Look at the sample sizes: apart from one study of 64 people, every group was 20 or fewer.

The one study that found REM decreased is also the one with the largest single dose. At 0.7 to 1.4 mg per kg, a 70 kg adult in that experiment would have received roughly 49 to 98 mg of oral THC in one go.[1] For a sense of how that compares with what is on Canadian shelves, see how strong Canadian weed is.

What does "REM rebound" mean?

REM rebound describes REM sleep returning more quickly and taking up more of the night after a period in which it was reduced. During cannabis abstinence, both shifts have been measured: REM arrived sooner and occupied a larger share of the night.[3]

Two separate measurements sit under the label:

  • REM latency: minutes from falling asleep to the first REM period. Shorter latency means REM arrives sooner.
  • REM percentage: the share of total sleep spent in REM.

On the withdrawal side, shorter REM latency is the most consistent finding. REM time is not consistent: one 2008 study found no change, a 2010 study found it went down, and a study of adolescents found no effect on REM.[4]

Here is the part most explanations skip. "REM rebound" is a description of measured shifts, not a demonstrated cause of what you feel. The step from "REM came sooner" to "that is why my dreams are so intense" is reasonable but inferred, because no one measured the dreams themselves.

How common are strange dreams during cannabis withdrawal?

About a third on average (34.4% pooled across 11 studies), but individual studies found anywhere from 5.1% to 78.0%. The highest rates come from studies of daily, dependent users; the lowest from general-population surveys.[4]

That fifteen-fold spread is why a single headline percentage is misleading. Three things drive it:

  • Who was studied. The 78% figure comes from dependent daily users; the 5.1% figure from a past-year survey of a broader population.
  • What was asked. One study asked about "sleeping difficulty or strange dreams" together; others asked about "nightmares", "strange dreams" or "unpleasant dreams". Those are not the same question.
  • When it was asked. Daily ratings during abstinence catch more than a single survey looking back over a year.
StudyMeasureReported rate
Budney 2003Sleep inventory and withdrawal checklist78%, "consistently elevated" (onset day 2, peak day 9)
Arendt 2007"Sleeping difficulty or strange dreams"72.2% mild, 55.6% moderate, 27.8% severe
Budney 1999Marijuana Withdrawal Checklist50% mild, 37% moderate, 20% severe
Vandrey 2005Marijuana Withdrawal Checklist49% (26% mild, 15% moderate, 8% severe)
Allsop 2011Cannabis Withdrawal Scale, daily41% "nightmares", peaking at 1.5 days
Levin 2010Checklist of 8 sleep items20.1% "strange dreams"
Agrawal 2008Past-year survey, 4 items5.1% "unpleasant dreams"
Pooled, 11 studiesAll of the above and others34.4% (range 5.1% to 78.0%)

Source: Gates et al. 2016.[4] One more finding from that review puts the numbers in perspective: in one study, strange dreams were the most commonly reported symptom yet ranked only 10th most distressing out of 26. Common does not mean the worst part.

Do edibles and smoking affect dreams differently?

No study has compared them. The sleep-lab research used smoked cannabis or oral THC in separate experiments, never head to head, so any confident answer about edibles and dreams is unverified rather than established.[1]

Unverified is not the same as disproven. It simply has not been tested. For how edibles differ in timing generally, see how long an edible high lasts.

How reliable is what you read about weed and dreams?

Less reliable than its confidence suggests. The central claim, that THC suppresses REM, rests on six small studies, four of which found no effect. And no study in this literature measured a dream: dreams appear only as a self-reported checklist item.[1][4]

Most pages on this topic tell a neat two-part story: THC suppresses REM, so quitting brings it roaring back. The second half has lab support. The first half does not hold up as stated. Here is what to watch for:

  • "THC suppresses REM" stated as fact. The objective studies are split one up, one down, four no effect. Mixed, not settled, and not disproven either.[1]
  • Old, tiny, high-dose studies treated as typical. The 1972 to 1982 studies had 2 to 11 participants each, and the one REM decrease came at 0.7 to 1.4 mg/kg of oral THC.[1][2]
  • "REM rebound" used as an explanation. It labels measured shifts in REM latency and percentage. The link to how a dream feels is inferred.[3]
  • One prevalence number with no range. 34.4% means little without 5.1% to 78.0% beside it.[4]
  • No comparison group. The key lab study did not include people who don't use cannabis.[3]
  • Study quality is modest. The withdrawal studies were rated "low to moderate" in quality (32.4% to 80.0%), and the administration studies averaged 42.6%.[1][4]

The meta-analysis adds one more steady point: stopping cannabis was consistently associated with sleep disturbance, even though using it showed no consistent effect on sleep stages.[2] The dreams are the solid part. The story told about them is the shaky part.

FAQ

Does weed stop you from dreaming?
That has not been shown. If you remember fewer dreams while using regularly, no study reviewed here measured dream recall during use, so the reason is unknown. Across six studies that measured sleep objectively, one found REM increased, one found it decreased and four found no effect, so whether cannabis changes REM sleep itself remains unsettled.
Why are my dreams so vivid after quitting weed?
Strange and vivid dreams are among the best-documented experiences after stopping cannabis. Pooled across 11 studies, an average of 34.4% of people reported them, with a range of 5.1% to 78.0%. In a sleep lab, REM sleep arrived sooner after falling asleep and took up a larger share of the night during abstinence.
How long do weed dreams last?
In the one study that tracked dreams day by day, strange dreams began around day 2 of abstinence and peaked around day 9. A separate study placed the peak for nightmares at around 1.5 days. Reports of strange dreams lasting up to seven weeks have been observed in some people, so timelines vary widely.
What is REM rebound?
REM rebound describes rapid eye movement sleep returning sooner and taking up more of the night after a period in which it was reduced. During cannabis abstinence, both of those shifts have been recorded in a sleep laboratory. The connection between them and how intense a dream feels is inferred rather than measured.
Does THC really suppress REM sleep?
Less clearly than most articles suggest. Reviewing six studies that used objective sleep measures, Gates and colleagues found REM increased in one study, decreased in one and showed no effect in four. A 2025 meta-analysis found no consistent effect on sleep staging. The evidence is mixed and not established, rather than disproven.
Do edibles affect dreams differently than smoking?
No study has compared the two routes for their effect on dreams. The sleep laboratory research used either smoked cannabis or oral THC in separate experiments, never head to head. Any confident answer about whether edibles change dreams differently is unverified rather than established, and is best treated that way until someone tests it.
Are nightmares part of quitting cannabis?
Nightmares and strange dreams are reported by people stopping cannabis, and in one study they were the most commonly reported symptom. That same study ranked them only tenth most distressing out of 26. Rates vary widely between studies, from about 5% in a general survey to 78% among dependent daily users.
Will my dreams go back to normal?
The studies that followed people over time recorded dream reports falling from their early peak, though none measured dreaming directly and none tracked people indefinitely. Reported durations differ a lot between studies. Sleep disruption that persists or affects daily life is worth raising with a clinician rather than researching online.

References

  1. Gates PJ, Albertella L, Copeland J. The effects of cannabinoid administration on sleep: a systematic review of human studies. Sleep Medicine Reviews 2014;18:477-487. doi.org
  2. Velzeboer R, Malas A, Wei S, Berger R, Parmar V, Lai WWK. Cannabis and sleep architecture: a systematic review and meta-analysis. Sleep Medicine Reviews 2025;84:102164. doi.org
  3. 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. doi.org
  4. Gates P, Albertella L, Copeland J. Cannabis withdrawal and sleep: a systematic review of human studies. Substance Abuse 2016;37(1):255-269. doi.org

For adults 19+. General information, not medical advice. In an emergency, call 9-1-1.