Do Sleep Supplements Work? Trial Evidence Reviewed

The sleep supplement market is worth billions and is built on a specific rhetorical move: taking a nutrient that participates in sleep biology, then implying that supplementing it improves sleep. Those are different claims, and the trials usually support only the first.

This is a category-by-category review of what is actually known, with attention to how big the effects are and who paid for the studies.

Table of Contents

How to read a sleep supplement trial

Three things separate a meaningful result from a press release:

Effect size, not p-value. A study can show a statistically significant improvement of four minutes in sleep latency. That is real and useless. Look for changes measured in tens of minutes, or in validated questionnaire scores with a known minimal clinically important difference.

Objective versus subjective outcome. Polysomnography and actigraphy measure sleep. Questionnaires measure how people feel about their sleep. Both matter, but they diverge, and supplement trials lean heavily on the subjective.

Who funded it. A substantial share of positive supplement trials are funded by manufacturers or by trade associations. That does not automatically invalidate them, but it should adjust your confidence, especially when the effect disappears in independent replication.

The categories

Melatonin — real effect, wrong dose on shelves. This is the one supplement with robust evidence for a specific indication: shifting circadian timing. For jet lag, the evidence is strong enough that it is recommended in clinical guidelines. For sleep onset in delayed sleep phase, also good. For general insomnia, the effect on sleep latency is modest — roughly 7 to 12 minutes in meta-analyses — which is real but less than people expect. The dose problem is covered in our melatonin dosage guide: trials used 0.3 to 3 mg, shelves sell 5 and 10 mg.

Magnesium — corrects a deficiency, does not sedate. Trials in older adults with low intake show modest benefit. Trials in people with adequate status show little. This is what you would expect from a nutrient. Details in our magnesium guide.

Valerian — weak, inconsistent, dose-dependent. The best-studied herb, with the most contradictory results. Some trials show improved sleep quality; others show no difference from placebo. A 2020 systematic review concluded the evidence was insufficient to recommend it. Effects, where present, take one to two weeks. Covered in the sleep tea guide.

Chamomile — small effect, excellent safety. A handful of small trials suggest modest improvements in sleep quality, particularly in postpartum women and older adults. Low risk, low cost.

Glycine — promising, under-studied. Glycine is an inhibitory neurotransmitter and a common component of collagen. Several small Japanese trials found that 3 g before bed improved subjective sleep quality and reduced core temperature, with better results in people with poor sleep. The mechanism is plausible (peripheral vasodilation and core temperature drop), the effect size is modest, and the studies are small. It is one of the more interesting candidates.

L-theanine — evidence for calm, not for sleep. An amino acid in tea. Trials show reduced subjective tension and improved relaxation at 100 to 200 mg. Effects on objective sleep measures are inconsistent. It is likely useful for the “wired but tired” state rather than for sleep itself.

Ashwagandha — industry-funded, effect uncertain. Several trials report improved sleep quality and reduced cortisol. Most are small, short, and manufacturer-funded, and ashwagandha has documented liver toxicity in a small number of cases. If you take it, get liver function checked if you develop symptoms, and do not combine it with other hepatotoxic substances.

GABA, 5-HTP, tryptophan — mechanism stories without good trials. GABA taken orally does not reliably cross the blood-brain barrier, which does not stop it being sold. 5-HTP is a serotonin precursor with cardiac and serotonergic interaction risks; tryptophan has modest evidence and a history worth knowing about. Neither is a first choice.

Combination products — untestable by design. A capsule containing melatonin, magnesium, L-theanine, glycine, valerian, and ashwagandha cannot be evaluated, because no trial can attribute the effect. If a combination works for you, you will never know which part to keep, and you will pay for six ingredients to get one.

Supplement	Evidence strength	Typical effect	Main caveat
Melatonin	Strong for circadian timing; modest for sleep onset	7–12 min faster onset	Dose on shelves is 5–10× the tested dose
Magnesium	Modest, only if intake is low	Small quality improvement	Label states compound, not elemental
Valerian	Weak, inconsistent	Variable	Takes 1–2 weeks; smells terrible
Chamomile	Weak but consistent	Small	Very safe
Glycine	Emerging	Small–moderate subjective	3 g needed; few trials
L-theanine	Moderate for relaxation	Calm, not sedation	100–200 mg; not a hypnotic
Ashwagandha	Weak, conflicted	Reported improvement	Liver risk; industry funding
Combinations	Not assessable	Unknown	Cannot attribute effect

What the supplements are competing with

The comparison that matters is not supplement versus nothing. It is supplement versus the things with much larger effect sizes:

CBT-I (cognitive behavioural therapy for insomnia) is first-line treatment for chronic insomnia in every major guideline, with effect sizes substantially larger than any supplement, and durable benefit after treatment ends.
Fixed wake time is the single most powerful behavioural lever, because it anchors the circadian clock more reliably than bedtime does.
Morning bright light phase-advances the clock and improves sleep onset at night. Covered in our light therapy guide.
Bedroom temperature. Core temperature has to fall for sleep to start; a room above about 19 °C makes that harder. See our guide to the 65 °F rule.
Caffeine timing. A 3 p.m. coffee still has a measurable effect at midnight in sensitive people.
Alcohol. It reduces sleep latency and then fragments the second half of the night. It is the most common self-treatment that makes the problem worse.

If you have tried none of these and are buying a supplement, the ordering is wrong.

If you do want to try something

A sensible sequence:

Fix the floor first — fixed wake time, caffeine cutoff, dark and cool room — for two weeks.
Melatonin at 0.5 mg if the problem is falling asleep or jet lag, taken at a consistent time.
Magnesium at 200 mg elemental in the evening if your diet is low in it.
Glycine at 3 g if you want to try something with an interesting mechanism and low risk.
One thing at a time, for two weeks each. Changing three things at once means learning nothing.
Stop what does not work. The supplement aisle relies on people continuing to buy things that did not help.

Safety, briefly

Supplements are not risk-free. The recurring issues are contamination (melatonin products with serotonin, herbal products with undeclared pharmaceuticals), mislabeling (content far from the claim), interactions (St John’s wort, valerian with sedatives, 5-HTP with antidepressants), and liver injury (ashwagandha, kava, some green tea extracts).

Look for USP, NSF, or Informed Choice verification. Buy single-ingredient products. Tell your pharmacist what you take, especially if you take anything acting on the central nervous system.

How we assessed this

Nothing on this page is a clinical recommendation. For each product we check four things: whether the dose in the product matches the dose used in the trials its own marketing cites; whether an independent certificate of analysis is published for the batch; what is in the “other ingredients” line; and how the dose sits against the upper limit set by a recognised authority.

What we cannot verify: we do not assay products ourselves. A supplement can pass every screen above and still fail a potency test, which is exactly why third-party certification carries so much weight in our ranking and why an uncertified product with an attractive label ranks below a certified one with a dull label.

One more limit worth naming: supplements are the category where marketing runs furthest ahead of evidence. If a claim on a bottle sounds like a drug claim, treat that as a reason to look harder at the certificate, not at the bottle. Anyone who is pregnant, on prescription medication, or treating diagnosed insomnia should speak to a pharmacist or doctor before buying.

The rules behind this section — how picks are made, what we disclose, and how corrections are handled — are written down in our editorial policy. If a pick here looks wrong to you, that page tells you how to tell us.

FAQ

Is there a supplement that reliably puts you to sleep? No. Melatonin shifts timing; it does not sedate. Nothing over the counter approaches the effect of a prescription hypnotic — which is also not first-line for chronic insomnia.

Do supplements help if my insomnia is stress-related? Less than behavioural treatment does. The evidence for CBT-I in stress-related insomnia is much stronger.

Can I take melatonin and magnesium together? Yes, at reasonable doses. Many people do. Watch for next-day grogginess, and remember magnesium has an upper limit of 350 mg supplemental per day.

Is it safe to take a sleep supplement every night? Melatonin short-term is well studied; long-term nightly use is common but less characterised. Herbal supplements are not intended for indefinite use.

What if nothing works? That is a signal to see a clinician, particularly if you snore heavily, wake unrefreshed, or fall asleep during the day — those point to sleep apnoea, which no supplement addresses. Our guide to a partner’s snoring covers when snoring is worth investigating.