Best Supplements for Sleep: Only One Is Supported, and It Buys Minutes
Melatonin is the most-studied thing on this list, and in a 2013 pooling of trials it got people to sleep about seven minutes faster; in adults alone, a 2022 pooling found about two. When a US team measured 25 melatonin gummies in 2023, 22 of them were more than 10% off their label, most of them over it. [the full verdict]
We have checked 10 of the things sold for sleep, one at a time, reading the trials rather than the packaging.
Exactly one of them is Supported for sleep itself. It is the one measured in minutes. Everything else on this page is either unsettled, settled for a different outcome entirely, or did not hold up.
Nothing here is for sale and nothing links to a shop. Every row goes to the verdict it came from, so you can check the working.
| What | Verdict | What we measured |
| Melatonin About seven minutes faster to sleep in a 2013 pooling, about two in adults alone. In a 2023 test, 22 of 25 US gummies were more than 10% off their label. | supported | 7 min faster to fall asleep in a 2013 pooling of adults and children; about 2 min in adults alone in 2022 |
| What | Verdict | What we measured |
| Magnesium It is a staple of sleep lists. What the evidence actually settles is that it lowers blood pressure a little, mostly when it is high. It does not stop older adults’ night leg cramps either. | supported | 2.0 mmHg off systolic blood pressure, across 34 double-blind trials |
| What | Verdict | What we measured |
| Mouth taping The best result came from 20 people with mild sleep apnea, blocked noses treated first. We found no test of taping over a nose that stays blocked. | preliminary | 20 people in the best result, all with mild sleep apnea, any blocked nose treated first |
| Blue-light blocking The mechanism is real. The link onward to recovery and gains is not. | preliminary | 1 in 5 studies finding sleep quality decreased, in the cited review |
| Putting the phone down 63 studies tie feeling hooked on your phone to bad sleep. Tested, the pull runs both ways; phone-free bedtimes helped in small trials. | preliminary | 63 studies — every one of them observational |
| Ashwagandha Sleep is its best-supported claim, on five small trials. Cortisol drops; whether people feel less stressed is contested, and most trials had a company’s money, product or staff. | preliminary | 0.40 p-value for perceived stress in one of four reviews; three found a drop |
| Box breathing Mood lifted a little after each session, as much as with two other breathing patterns. Tried over four weeks of daily life, it brought no clear drop in stress. | preliminary | 21 people in the box-breathing arm of the month-long trial |
| What | Verdict | What we measured |
| Sleep hygiene, on its own The insomnia guideline advises doctors against it as a standalone treatment: the only treatment it advises against. | unsupported | 1.01 what sleep hygiene adds inside CBT-I. 1.00 is nothing |
| Magnesium glycinate specifically The absorption trial behind “the most bioavailable form” was twelve people with surgically shortened bowels, and it came out even. Two later single-dose studies, each from the maker of a rival product, did not put it ahead either. | unsupported | 4 patients the absorption advantage came from |
| Tart cherry juice The famous 84-minute result came from eight people, and ten trials pooled show no clear effect on sleep. Its better evidence is about recovery after exercise, and that is not settled either. | unsupported | 8 people the 84-minute result came from |
Established means multiple good studies agree: treat it as settled. Supported means the evidence is consistent, with known limits. Preliminary means the evidence is early, small, conflicting or from animals only: not enough to act on alone. Unsupported means we looked and the evidence does not back the claim: either it was tested and did not hold up, or nobody has tested it and nothing that was tested supports it.
Nothing on this page reaches Established, and that is worth sitting with. Sleep is among the most-studied things a person can measure, the market for fixing it is enormous, and the best-evidenced product in it moves the number by minutes: about seven in a 2013 pooling, about two in adults alone. The gap between how confident the packaging sounds and how confident the evidence is may be wider here than anywhere else we have looked.
The same corpus, asked other ways: for weight loss, where nothing reaches Established for weight loss either, for muscle growth, for men and for women, for gut health, and for recovery, and for sweeteners, where every claim is a worry and the ratings run backwards.
The four ratings are the same ones used on every verdict here, and we wrote down how we read a study so you can see what earns each one.