Relaxation techniques for sleep, ranked by the evidence

Every sleep article hands you the same list: breathe slowly, tense and release your muscles, picture a beach, be grateful, meditate. Almost none of them say which items on that list have been tested, or how well. So here are ten techniques, ranked by evidence. The short answer: progressive muscle relaxation wins, with 31 randomized trials behind it and a 3.79-point improvement on the Pittsburgh Sleep Quality Index. Slow breathing and imagery follow on solid ground, and the trendier entries further down run mostly on anecdote.

Can relaxing at bedtime backfire?

A network meta-analysis of 241 trials and 31,452 people pulled apart which components of insomnia therapy do the work. Cognitive restructuring, sleep restriction, and stimulus control all helped. Relaxation was the one component the authors flagged as potentially counterproductive (Furukawa 2024).

That deserves context in both directions. The estimate was uncertain: the confidence interval still crossed no-effect, and the authors wrote “potentially,” not “is.” But take the reasoning seriously: relaxing at bedtime can encourage you to lie in bed awake longer, which works directly against sleep restriction and stimulus control.

So the frame for everything below is this. These techniques are for the specific problem of a body and mind too wound up to let go. None of them is a treatment for chronic insomnia, and if you’re using one to justify another 45 minutes of lying there, it’s costing you.

Tier 1: which techniques are well tested?

Progressive muscle relaxation. The biggest surprise of this research. A 2026 meta-analysis of 31 randomized trials and 2,277 adults found PMR improved sleep quality with a PSQI difference of −3.79, a larger effect than the entire Cochrane review on bedtime music (Donato 2026). It comes with a serious asterisk about heterogeneity, which we unpack in the full post on PMR. But nothing else on this list has that much trial evidence behind it.

Slow breathing. The physiology is the best-established thing here: breathing under about 10 breaths a minute reliably raises heart-rate variability and vagal activity, shifts EEG toward alpha, and makes people report feeling calmer (Zaccaro 2018). The sleep evidence is thinner and there’s one large, well-designed null result that everyone selling breathwork ignores; the breathing post covers both.

Guided imagery. Forty-one people with insomnia were told at bedtime to either distract themselves with vivid imagery, distract themselves generally, or do nothing. The imagery group fell asleep faster and reported less frequent and less distressing pre-sleep thinking. Plain distraction didn’t do it (Harvey 2002).

That last detail is the useful one. “Stop thinking” fails; “think about this instead, in detail” works. It’s also the mechanism behind sleep stories, which are imagery-distraction tasks somebody else built for you.

Tier 2: what has real evidence but gets oversold?

Meditation. More studied than anything else on this list, and the results are more equivocal than you’d guess. Against inactive or attention-matched comparisons it does help sleep; against actual evidence-based sleep treatments, the pooled effect is essentially zero (Rusch 2019). And a much larger review of 47 trials found low or insufficient evidence of any effect on sleep at all (Goyal 2014).

There’s also a mismatch nobody mentions: the trials tested six- to eight-week courses with a live teacher and homework, not a twenty-minute track at bedtime. The meditation post is the long version, because this is the technique most likely to be mis-sold to you.

Yoga nidra, lately rebranded as NSDR. The strongest single trial randomized 41 people with chronic insomnia to yoga nidra or CBT and found it held its own on polysomnography, including a shift toward deep sleep (Datta 2021). Pool the randomized trials, though, and neither sleep quality nor insomnia severity reaches significance; the reviewers rate the certainty of evidence very low (Singh 2026). What the yoga nidra research shows goes further, including why we’d still rank it above its evidence grade.

Hypnosis and hypnotic suggestion. Better than its reputation. In a placebo-controlled crossover experiment, 14 highly suggestible women who listened to a hypnotic suggestion to “sleep deeper” before a nap got 81% more slow-wave sleep and 67% less time awake (Cordi 2014). Startling numbers from real polysomnography, and from fourteen people: that figure is the first of five experiments the paper reports across 70 women in total. What happened in the others is the subject of a later section.

The caveat is enormous. A systematic review of 24 papers called hypnosis promising for sleep with few adverse events, but limited by small samples and low methodological quality; 58% of the included studies reported a benefit and 29% reported none (Chamine 2018).

Tier 3: what’s interesting but unproven?

The body scan. The technique inside every other technique: it sits inside MBSR, inside yoga nidra, and inside half the guided sleep meditations on your phone, and it has never been tested by itself. The nearest direct evidence is small and qualitative, and it’s a point in its favor: insomnia patients finishing an eight-week mindfulness course singled out the body scan as the tool that got them to sleep faster (Hubbling 2014). The body scan on its own covers the borrowed evidence, and who should point their attention elsewhere.

The military sleep method. The viral two-minute technique with the 96% success rate. The number is a track coach’s 1981 account of Navy pilots, not a study, and no trial has tested the method since. Its four steps are lifted from techniques that do have evidence, muscle relaxation and imagery among them, which is why it sits here rather than lower. The military method, step by step sorts the tested parts from the marketing.

Gratitude. The foundational finding is correlational: across 401 people, trait gratitude predicted better sleep quality, longer sleep, and shorter sleep latency, with the link running through more positive and fewer negative pre-sleep thoughts (Wood 2009). That’s a plausible mechanism, but a cross-sectional study can’t tell you that practicing gratitude causes better sleep. The experimental work is smaller: a two-week gratitude intervention in 119 young women reported improved sleep quality alongside gains in wellbeing and optimism (Jackowska 2016). Promising, thin, worth three minutes of your evening rather than a system.

ASMR. The biggest gap on this list between how popular something is and how much is known about it. The core findings are that 82% of ASMR-experiencers use it to help them sleep (Barratt 2015), and that people who experience ASMR show heart-rate drops averaging about 3 bpm while watching ASMR videos, though their skin conductance rises at the same time, so it isn’t simply calming (Poerio 2018). Both use self-selected samples of people who already know it works for them. Sleep-specific experimental work is small and some of it hasn’t cleared peer review. If it works for you, use it; just know the science is behind the practice, not ahead of it.

Why does the same technique work for your friend and not for you?

Here’s the thread that runs under all of this, and it’s the most useful thing in the post.

Remember the hypnosis result, slow-wave sleep up 81%. That happened in highly suggestible participants. In the low-suggestibility experiments, the same recording did nothing useful, and the instruction to “sleep deeper” decreased their slow-wave sleep, to about two-thirds of their own control-night level.

Now put it next to a different study by the same lead researcher. Twenty-seven women listened to relaxing music before a nap; the ones who got more slow-wave sleep were the participants who scored low on suggestibility (Cordi 2019).

Read together: suggestibility isn’t a flaw in these studies, it’s a variable that decides which intervention you respond to. Techniques that work through suggestion (hypnosis, “your legs are getting heavy,” visualization scripts) need you to be the kind of person who takes a suggestion. Techniques that work mechanically (slow breathing moving your vagal tone, muscle tension physically releasing, sound masking a noisy street) don’t care what you believe.

Which is why “just try meditation” is bad advice and “try three of these for a week each” is good advice. You’re not looking for the technique with the best average. You’re looking for the one that works on you.

So which technique should you try first?

Start with the two that are least dependent on belief: slow breathing and progressive muscle relaxation. Give each about a week; for the breathing half, our breathing pacer keeps the rhythm while you learn it. If neither lands, try an imagery-based approach (a story or a visualization), which is where the suggestible crowd tends to do well. Keep whichever one you stop noticing you’re doing.

And keep the Furukawa warning in view: if you’re twenty minutes in and still awake and annoyed, get up and do something dull in dim light. No technique on this list beats not lying there.

One more fork in the road. Everything above treats the problem as a wound-up body and mind. If what’s keeping you up is the room (a snoring partner, a loud street, a hotel), no relaxation technique reaches it; that job belongs to steady sound, and the noise-color explainer covers how masking works and how to pick.

If you want the guided version

Slumber is built out of exactly these techniques: guided meditations, breathing and muscle-relaxation exercises, and calming imagery, in a library of 2,200+ stories and meditations with new ones added daily. Trying three techniques for a week each is a lot easier when each one is a track you press play on.

---

Sources: Furukawa et al. 2024 · Donato et al. 2026 · Zaccaro et al. 2018 · Harvey & Payne 2002 · Rusch et al. 2019 · Goyal et al. 2014 · Datta et al. 2021 · Singh et al. 2026 · Hubbling et al. 2014 · Cordi et al. 2014 · Chamine et al. 2018 · Wood et al. 2009 · Jackowska et al. 2016 · Barratt & Davis 2015 · Poerio et al. 2018 · Cordi et al. 2019. This is general information, not medical advice — persistent insomnia is worth taking to a doctor.