The one subject on this site with laboratory evidence behind it — and, for the same
reason, the one where the honest success rates are lower than anybody advertises.
Saunders et al. 2016 · 34 studies · 24,282 participants
How common it actually is
How many people have lucid dreams, and how often
More than half of people have had at least one lucid dream (55%), but fewer than one in four has them every month (23%) and about one in nine has two or more a month (11%).
So it is common to have had one, and uncommon to have them regularly. Any page telling
you this is a nightly skill is describing a small minority. Any technique that works on
one night in ten or more is doing something real, and it is still not the switch these
methods are usually sold as.
18 pages · printable
The 2027 Sky Calendar — free PDF
Every retrograde, eclipse, new and full moon of 2027 to the minute,
already converted to your city. A month to a page, ready to print.
Nobody else calculates this
When to set the alarm
Every guide says “wake up after about five hours”. Five hours from what — getting into
bed, or falling asleep? And why five? The answer is the 90-minute sleep cycle: you want
to wake between cycles in the last third of the night, where REM periods run 40
minutes instead of 10. Here are your actual clock times.
The techniques, with what the evidence says
Success rates are what has been measured, not what is promised. Where nothing has been
measured, that is stated instead of filled in.
Where each technique sits in the day and the night
Reality testing is the only daytime method. WBTB is the alarm itself, and MILD, SSILD and WILD are what you do with that waking before the long REM periods at the end of the night.
Stephen LaBerge, 1980 · success: 17% of nights, up to 46%
The best-evidenced technique. In Aspy et al. (2017), MILD combined with reality testing and WBTB produced lucid dreams on 17% of nights overall, rising to 46% among participants who fell asleep within five minutes of the rehearsal.
How to run it. On waking from a dream, rehearse it, then repeat the intention — "next time I am dreaming, I will remember I am dreaming" — while visualizing re-entering that dream, and fall back asleep holding the intention.
Needs a night waking — pair it with the alarm times above.
Folk technique, first tested by LaBerge's team in 1994 · success: Not measured alone — it boosts other methods
Rarely used alone — it raises the odds of whatever else you are doing, because it lands you in the REM-dense final third of the night. Aspy (2017) treats it as a component rather than a standalone method.
How to run it. Set an alarm for 4.5 to 6 hours after falling asleep, stay awake 10 to 30 minutes, then return to bed applying MILD. The alarm-time calculator works out the clock times for you.
Needs a night waking — pair it with the alarm times above.
Documented by LaBerge, 1985 · success: Not reliably measured
Entering a dream directly from waking. The hardest to do and the least measured — and the one most likely to produce sleep paralysis, since you are deliberately staying conscious while the body locks down.
How to run it. From a WBTB waking, lie still and hold attention on a single point — the breath, a counted phrase — through the hypnagogic imagery, without moving, until a dream scene forms around you.
Needs a night waking — pair it with the alarm times above.
Community-developed, around 2011 · success: About as effective as MILD
Developed on online forums and later tested: in the 2020 International Lucid Dream Induction Study, with 355 participants, SSILD and MILD were similarly effective, and a hybrid of the two showed no extra benefit.
How to run it. After a WBTB waking, cycle attention through sight, hearing and touch — a few seconds each, several rounds — then let yourself fall asleep.
Needs a night waking — pair it with the alarm times above.
Paul Tholey, 1983 · success: Weak alone, useful combined
On its own the effect is small. Its value is in feeding MILD: the habit of questioning your state carries into the dream, where the check fails and the failure is the cue.
How to run it. Several times a day, genuinely ask whether you are dreaming and test it — read text twice, look at your hands, check a clock. In dreams text tends to shift and clocks misbehave.
Runs during the day, no night waking needed.
Why this one is different from the rest of the site
Almost everything else here is tradition — meanings people wrote down, with no mechanism
and no evidence. Lucid dreaming is not. In 1981 Stephen LaBerge had sleeping participants
signal awareness with pre-agreed eye movements, recorded on EOG, while polysomnography
confirmed REM sleep. The signals came through. That is a verified phenomenon, not an
interpretation.
Which is exactly why the numbers on this page are lower than the ones you will read
elsewhere. Once something is measurable, it stops being possible to promise.
Start with recall, not with technique
The common failure is jumping straight to WILD. If you do not remember ordinary dreams,
you will not remember lucid ones either, and you will conclude the technique failed when
recall was the problem. Keep a dream journal for two weeks first — write on waking,
before moving — and only then add reality checks and MILD. Dream recall improves fast
with journalling, and it is the input every other method depends on.
Sleep paralysis, plainly
Techniques that hold consciousness through sleep onset make sleep paralysis more likely.
What it is: you notice the atonia that normally accompanies REM and that you normally sleep through.
How long: it is harmless and ends on its own, usually within a minute or two.
Why it matters: it is genuinely unpleasant, and it is the main reason to be cautious with WILD rather than with the gentler methods.
Questions about lucid dreaming
What is a lucid dream?
A dream in which you know you are dreaming while it is happening. It is not a belief or an interpretation — it has been verified in the laboratory: in LaBerge’s 1981 work, sleeping participants signaled awareness with pre-agreed eye movements recorded on EOG while polysomnography confirmed they were in REM sleep. That makes lucid dreaming one of very few subjects on this site with actual experimental support.
How common are lucid dreams?
55% of people report at least one in their lifetime, 23% have them monthly or more often, and about 11% report two or more a month. Those figures come from Saunders et al. (2016), a meta-analysis of 34 studies covering 24,282 participants — not from a survey somebody ran on a forum.
How do I lucid dream tonight?
Realistically you probably will not. The best-evidenced method, MILD combined with wake-back-to-bed, produced lucid dreams on 17% of nights in Aspy et al. (2017) — rising to 46% for people who fell asleep within five minutes of the rehearsal. Those are the honest numbers, and they mean it is a practice rather than a trick.
When should I set my alarm for WBTB?
Between 4.5 and 6 hours after you fall asleep, which lands you between sleep cycles in the part of the night where REM periods are longest. The exact clock time depends on your bedtime — the calculator on this page works it out from a 90-minute cycle plus about 15 minutes to fall asleep.
Is lucid dreaming dangerous?
For most people it is not, but two things are worth knowing. Techniques that keep you conscious into sleep onset — WILD especially — raise the chance of sleep paralysis, which is harmless but frightening. And any method built on waking yourself at night costs you sleep continuity, which matters more than the dreams do.
Statements of fact on this page were checked against these sources in October 2026.
How we choose and check sources: editorial policy.
Cite this page
AstroHot. “Lucid Dreaming: What the Research Shows + WBTB Calculator.” Accessed November 2, 2026. https://astrohot.com/dream-dictionary/lucid-dreaming/
For Wikipedia{{cite web |title=Lucid Dreaming: What the Research Shows + WBTB Calculator |url=https://astrohot.com/dream-dictionary/lucid-dreaming/ |website=AstroHot |access-date=2026-11-02}}
Free · PDF · 18 pages
The 2027 Sky Calendar
Every retrograde, eclipse, new and full moon of 2027 — to the minute, printed in your own time zone.
