Natural Sleep Aids: What the Evidence and Safety Data Show

A product described as a natural sleep aid is not automatically effective, standardized, non-habit-forming, free of interactions, or safer than a medicine. In the United States, the Food and Drug Administration does not approve dietary supplements for safety, effectiveness, or labeling before they are sold. Manufacturers carry the initial responsibility for those requirements, and FDA oversight is largely postmarket ¹.

Some products have a plausible, limited role for a specific sleep problem. That is different from proving that a whole category treats insomnia. Melatonin is the clearest example: it can act as a timing signal for selected circadian problems, but it is not a general sedative and major guidelines do not recommend it as a routine treatment for chronic insomnia. Evidence for magnesium, valerian, chamomile, passionflower, lavender aromatherapy, tart cherry, glycine, and CBD is limited, inconsistent, or tied to a particular formulation and population ². There is no evidence-based “best” natural sleep aid or universal dose, timing, or course for every reader.

Start with the sleep problem, not the product

“Help me sleep” can mean several different things:

• Occasional sleep difficulty: A few disrupted nights during travel, stress, illness, or a schedule change do not establish a sleep disorder. A product trial should still have a clear target, such as time to fall asleep or next-day alertness.
• A circadian timing problem: Someone may sleep normally on a late schedule but struggle to sleep at the time work or school requires. Correctly timed melatonin may be considered for some diagnosed circadian rhythm sleep-wake disorders, but the useful timing depends on the disorder and the person’s biological clock ³.
• Chronic insomnia: Persistent difficulty falling asleep, staying asleep, or returning to sleep despite adequate opportunity needs an insomnia assessment. Cognitive behavioral therapy for insomnia, or CBT-I, is the recommended initial treatment for adults, not a rotating stack of supplements ⁴.
• Another diagnosed or suspected disorder: A supplement does not treat obstructive sleep apnea, narcolepsy, restless legs syndrome, a medication effect, or another cause simply because it causes drowsiness. The underlying problem needs its own evaluation and treatment.

This distinction matters because a trial in healthy adults with mild sleep complaints cannot establish treatment for diagnosed insomnia, and an insomnia trial cannot establish treatment for a circadian disorder.

Natural sleep aids at a glance

Aid	What has actually been studied	Evidence-based reading
Melatonin	Selected circadian disorders, jet lag, shift-work-related sleep problems, and several insomnia outcomes	Useful in some timing-specific situations, but not a reliable general sedative or routine chronic-insomnia treatment 35
Magnesium	Different oral salts, mostly in small trials and selected groups, including older adults with insomnia	A 2021 review found low- to very-low-certainty evidence, which does not establish a benefit for the general population 6
Valerian	Variable root preparations and valerian combinations in chronic-insomnia trials	Results are inconsistent, products are not interchangeable, and an AASM guideline recommended against its use for chronic insomnia 5
Chamomile	Teas and extracts in varied populations, with sleep quality and insomnia measured differently	Some pooled sleep-quality findings are positive, but the single insomnia trial in a 2019 review did not improve insomnia severity 7
Passionflower	One-week tea use in healthy adults and short extract trials in adults with insomnia	Early signals are formulation-specific and too limited to establish a dependable insomnia treatment 8
Lavender aromatherapy	Inhaled lavender interventions, often measuring reported sleep quality	A recent review found a possible subjective benefit, but limited study quality and quantity prevent a firm treatment conclusion 9
Tart cherry	Juices, concentrates, powders, and combination products in small, short studies	Results are preliminary and cannot be transferred from one preparation, population, or blend to another 10
Glycine	A few small studies in healthy adults with sleep complaints or short-term sleep restriction	The sleep evidence has small samples and high risk of bias, so an effective regimen has not been established 11
CBD	Purified or defined CBD products, cannabinoid mixtures, and consumer products that differ substantially	A small 150 mg CBD pilot found no advantage over placebo for most sleep outcomes; it does not validate consumer gummies, oils, or blends 12

The table is not a ranking. A weak signal in a small trial does not become stronger because a product is popular, contains a nutrient found in food, has a traditional use, or has a plausible laboratory mechanism.

Melatonin is mainly a timing signal

The brain normally releases melatonin in relation to darkness and circadian timing. A supplement can shift that timing, which is why the hour it is taken may matter as much as the amount. The AASM circadian guideline suggests strategically timed melatonin for selected adults and children with specific circadian rhythm sleep-wake disorders, but its recommendations differ by diagnosis and population ³. A person who simply feels “wired at bedtime” may not have the same problem.

For chronic insomnia in adults, the AASM pharmacologic guideline suggested not using melatonin because the available trials did not provide strong enough evidence of clinically meaningful benefit ⁵. That conclusion does not mean melatonin never has a role. It means circadian use and chronic-insomnia treatment should not be treated as the same indication. Our detailed melatonin guide explains those distinctions.

There is no universal dose or clock time for every reader. The wrong timing can be unhelpful or move the circadian clock in an unintended direction. Formulation also matters: an immediate-release product and an extended-release product do not create the same exposure pattern.

Product content is another limitation. In a 2023 analysis of 25 melatonin gummies sold in the United States, 22 were inaccurately labeled. The measured melatonin ranged from 74% to 347% of the labeled amount among products containing detectable melatonin ¹³. This was a sample of gummies, not proof that every melatonin product is mislabeled, but it shows why a printed amount is not a guarantee.

Short-term use appears relatively safe for many adults, but long-term safety is not established. Reported effects include sleepiness, headache, dizziness, and nausea. People taking blood thinners or living with epilepsy need medical supervision, evidence in pregnancy and breastfeeding is inadequate, and melatonin can remain active longer in older adults. Children need clinician-guided decisions and secure storage because dosing and long-term developmental questions remain and accidental ingestions have caused serious outcomes ¹⁴.

Magnesium is a nutrient, not a proven sedative

Magnesium is essential to normal nerve and muscle function. That biological role, the magnesium content of foods, and an association between low intake and poor sleep do not prove that a supplement treats sleep difficulty.

A 2021 systematic review found only three randomized trials involving 151 older adults with insomnia. Magnesium shortened estimated sleep-onset time in the pooled analysis, but the trials had moderate-to-high risk of bias and the evidence was rated low to very low certainty. Total sleep time did not improve significantly ⁶. Those findings do not establish a benefit for younger adults, people without insomnia, or every magnesium salt.

A label should state the amount of elemental magnesium, not just the weight of the entire compound. Different salts have different absorption and gastrointestinal effects, but no form has been established as the best sleep treatment. High supplemental intakes can cause diarrhea, nausea, and abdominal cramping. Toxicity risk rises when kidney function is impaired, and magnesium can interfere with some antibiotics and osteoporosis medicines ¹⁵. A nutrient deficiency should be evaluated from the person’s diet, health conditions, medicines, and appropriate testing, not inferred from poor sleep alone.

Valerian has inconsistent insomnia evidence

Valerian studies have used different species, plant parts, extracts, doses, treatment periods, and combination products. That makes a positive result from one preparation difficult to apply to a tea, tincture, capsule, or blend with a different composition. In the trials reviewed for the AASM guideline, the evidence was not strong enough to support valerian for chronic insomnia ⁵.

Short-term use by most adults appears reasonably tolerated, but long-term safety is unknown. Reported effects include headache, stomach upset, mental dullness, uneasiness, vivid dreams, and rare liver injury reports, often involving multiple herbs. Withdrawal symptoms have also been reported after abrupt discontinuation following chronic use. Valerian should not be combined with alcohol or sedatives because its possible sleep-inducing effects may add to theirs ¹⁶.

Chamomile depends on the preparation and the outcome

Chamomile tea can be a pleasant caffeine-free drink, but a bedtime ritual is not the same as an insomnia treatment. A 2019 review pooled 12 randomized or quasi-randomized trials across anxiety and sleep questions. It found an improvement in pooled sleep-quality measures, yet the one trial that specifically assessed insomnia found no significant change in insomnia severity ⁷. Tea, capsules, and standardized extracts also deliver different compounds and amounts, so their findings should not be exchanged.

Chamomile in ordinary food or tea amounts is likely safe for many people, but allergy is a real boundary. Reactions are more likely in people allergic to ragweed, chrysanthemums, marigolds, or daisies, and severe reactions have occurred. Reported or theoretical interactions include warfarin, medicines metabolized by the liver, and sedatives. Pregnancy and breastfeeding safety data are limited ¹⁷. The Sleepiverse tea guide covers beverage-specific considerations such as caffeine, reflux, and nighttime urination.

Passionflower has small, short trials

Passionflower tea and extracts should be evaluated separately. A one-week crossover study in 41 healthy adults with mild fluctuations in sleep found a small improvement in self-rated sleep quality, but only 10 participants completed the laboratory sleep component ¹⁸. A later two-week trial in 110 adults with insomnia found a between-group improvement in total sleep time for a specific extract, while between-group differences in sleep efficiency and wake time after sleep onset were not significant ⁸.

These studies are signals for further research, not a universal dose or proof that any passionflower product treats insomnia. Passionflower can interact with anesthesia and other medicines used around surgery. NCCIH advises discussing its use if surgery is scheduled and avoiding it during pregnancy because it may cause uterine contractions. Breastfeeding safety is not established ¹⁹.

Lavender evidence applies to aromatherapy, not every lavender product

A 2026 systematic review of 11 randomized trials involving 628 adults found an improvement in sleep-quality scores with lavender essential-oil interventions, but the authors said the small number and quality of the included studies require better confirmation ⁹. These trials mostly support, at most, a possible change in reported sleep quality. They do not establish that lavender treats chronic insomnia, increases a particular sleep stage, or works through a claimed heart-rate, blood-pressure, or neurotransmitter mechanism.

Route matters. Inhaled lavender aromatherapy, a diluted topical product, lavender tea, and a purpose-made oral lavender preparation are different interventions. A bottle of concentrated essential oil sold for aromatherapy is not for ingestion. Aromatherapy can cause headache or coughing, and topical lavender can trigger a skin reaction. Possible sedative interactions and limited pregnancy and breastfeeding data also warrant caution ²⁰. Our essential-oil guide covers route-specific use in more detail.

Keep concentrated oils locked away from children. If someone swallows an essential oil, contact a local poison center or emergency service promptly for case-specific instructions ²¹. Oils should also be kept away from pets and never applied directly to them. Pet risk varies by species, oil, concentration, and exposure route; birds and pets with respiratory disease need particular care around diffusers. Contact a veterinarian or animal poison service after a suspected exposure ²².

Tart cherry findings are preliminary and product-specific

Tart cherry has been studied as juice, concentrate, powder, and part of multi-ingredient formulas. A 2025 systematic review found six clinical studies across different populations. Several reported an improvement in at least one sleep outcome, but the trials were small, brief, and heterogeneous. Some products contained other active ingredients, and not every objective outcome improved ¹⁰.

One small result, including the often-repeated increase in sleep duration from a study of older adults with insomnia, should not be converted into a promise for everyone or a serving recommendation. Melatonin or tryptophan content and anti-inflammatory mechanisms are hypotheses, not proof that cherries treat insomnia. Compare the exact preparation, serving, added ingredients, and population with the study before assuming the result applies.

Glycine remains an early research question

Glycine is an amino acid present in the body and in protein-containing foods. Its presence in food and its proposed effects on temperature regulation do not establish a sleep benefit from a supplement.

A 2024 systematic review found that the sleep-related human studies in healthy adults used small samples and had high risk of bias. Some studies reported better subjective sleep or next-day functioning, often after 3 grams before bed, but that studied amount is not a universal recommendation and long-term efficacy remains unclear ¹¹. There is not enough evidence to identify who is likely to benefit, whether the effect persists, or how glycine compares with established insomnia care.

Consumer CBD is not the FDA-approved CBD medicine

The only FDA-approved CBD product is a prescription medicine for specific seizure disorders. Approval for that product does not establish that an over-the-counter CBD oil, gummy, beverage, vape, or combination treats sleep problems. FDA has not evaluated consumer CBD products for a sleep indication, proper dosing, interactions, or consistent quality, and testing has found that some products do not contain the CBD amount claimed ²³.

Evidence specific to insomnia is limited. In a 2024 pilot trial, 30 adults with moderate-to-severe primary insomnia received either 150 mg of a defined sublingual CBD product or placebo nightly for two weeks. Insomnia severity, self-reported sleep onset, sleep efficiency, and wake time after sleep onset did not differ between groups. One objective sleep-efficiency measure and well-being improved, but most sleep outcomes did not ¹². This finding cannot be transferred to a lower-dose gummy, a CBD and THC mixture, or an unlabeled consumer product.

CBD can cause sleepiness or reduced alertness, diarrhea, appetite changes, and mood changes. It can injure the liver and change how other medicines work. Alcohol and medicines or products that slow brain activity can increase sedation and injury risk. Consumer products may also contain unexpected THC or contaminants ²³²⁴. FDA strongly advises against CBD during pregnancy or breastfeeding ²⁴. Anyone considering CBD should discuss the exact product with a clinician or pharmacist, especially if they take medicines or have liver disease.

How to judge a product label

A thoughtful purchase check cannot prove that a product works, but it can reduce avoidable uncertainty.

1. Prefer one identifiable ingredient. Look for the ingredient name, botanical species and plant part when relevant, formulation or extract, amount per serving, serving size, other ingredients, lot number, and manufacturer contact information.
2. Avoid hidden-dose blends. A U.S. proprietary-blend label may state only the total blend amount rather than the amount of each ingredient. The ingredients must be listed by weight, but the missing amounts make it difficult to compare the product with research or assess combined exposure ²⁵.
3. Treat certification as a quality check, not proof of benefit. Independent programs such as NSF or USP may help verify manufacturing, label identity, and contaminant limits. NIH’s Office of Dietary Supplements makes clear that these seals do not guarantee safety or effectiveness ²⁶.
4. Do not transfer evidence across forms. Tea is not an extract, aromatherapy oil is not an oral product, immediate-release is not extended-release, and a multi-ingredient gummy is not the isolated compound used in a trial.
5. Be skeptical of disease-treatment claims. “Clinically studied” may refer to an ingredient rather than the finished product. Natural, drug-free, non-habit-forming, detoxifying, and doctor-formulated are not evidence standards.

Combination “sleep blends” create two problems at once. They can hide the amount of each component, and they make it difficult to identify which ingredient caused a benefit, next-day impairment, allergic reaction, or drug interaction. Adding another product on top makes that uncertainty worse.

Check safety before the first use

Ask a clinician or pharmacist to review the exact label before use if any of these situations applies:

• You take prescription medicines, over-the-counter medicines, or other supplements. Interactions can change drug exposure or add sedation ¹.
• You use alcohol, cannabis, an opioid, an antihistamine, an anxiety medicine, a sleeping medicine, or another product that can reduce alertness ¹⁶²³.
• You have liver disease or impaired kidney function. CBD has a liver-injury risk, rare liver injury has been reported with valerian, and impaired kidneys increase magnesium-toxicity risk ²³¹⁶¹⁵.
• You have a plant, food, or product allergy. Chamomile and topical lavender have specific allergy and skin-reaction concerns ¹⁷²⁰.
• You are preparing for surgery. Supplements may interact with anesthesia or affect bleeding, blood pressure, or heart rate, so give the surgical team a complete product list and follow its stop or continue instructions ²⁷.
• You are pregnant, trying to become pregnant, or breastfeeding. Safety evidence is limited for many of these products, passionflower should be avoided during pregnancy, and FDA advises against CBD during pregnancy and breastfeeding ¹⁹²⁴.
• The intended user is a child or an older adult. These groups need product-specific guidance rather than an adult label copied across ages. For melatonin, long-term pediatric effects remain uncertain, and the supplement may remain active longer in older adults ¹⁴.

If a product causes sleepiness, dizziness, slowed thinking, poor coordination, or reduced alertness, do not drive, cycle in traffic, operate machinery, or do other safety-sensitive work. CBD in particular can make driving unsafe ²⁸.

A low-friction way to decide

1. Name the problem. Write down whether the target is sleep timing, time to fall asleep, awakenings, total sleep time, or next-day function. Do not use a supplement to mask snoring, gasping, breathing pauses, severe sleepiness, or a known untreated disorder.
2. Check fit and risk. Compare the exact formulation and population with the evidence, then review medicines, health conditions, pregnancy, surgery, allergies, and age with a clinician or pharmacist.
3. Choose one clearly labeled product only if it is appropriate. Avoid proprietary blends and do not start several aids together.
4. Define the outcome and review point before starting. Track the target outcome and next-day alertness, plus any adverse effects. A simple sleep diary is more useful than changing the plan after every difficult night.
5. Stop and reassess. Stop after a concerning reaction or meaningful next-day impairment. If the defined outcome does not improve by the agreed review point, reassess the diagnosis and plan instead of increasing, combining, or continuing indefinitely.

Persistent insomnia deserves evaluation and access to CBT-I, which has stronger evidence than these products as a class ⁴. Loud snoring, gasping, witnessed breathing pauses, or severe daytime sleepiness can point to sleep apnea or another disorder and should be assessed ². If sleepiness makes driving or hazardous work unsafe, stop the activity immediately and arrange safe transportation and prompt medical help rather than relying on a sleep aid, caffeine, or another quick fix ²⁸.