Melatonin

Overview

Melatonin is the hormone your pineal gland releases when the light goes down, and its job is to tell the rest of the body what time it is, not to knock you out. Taken at the right time, a dose small enough to mimic your own nighttime level shortens the time it takes to fall asleep and nudges a jet-lagged body clock. The catch: the trials that proved that used 0.3 to 1 mg, and most bottles on the shelf hold 3 to 10 mg.

What is Melatonin?

Key takeaways

• A 0.1 to 0.3 mg dose raises blood melatonin into the normal nighttime range, and 0.3 and 1 mg shortened time to fall asleep in MIT's sleep-lab trials with no morning hangover (Dollins 1994, PNAS; Zhdanova 1995, Clinical Pharmacology and Therapeutics).
• Pooled across 19 placebo-controlled trials in 1,683 people, melatonin cut time to fall asleep by about 7 minutes and added about 8 minutes of total sleep: real, consistent, and modest (Ferracioli-Oda 2013, PLoS One).
• In adults over 50 with poor sleep efficiency, 0.3 mg restored sleep efficiency to normal; 3 mg also helped but lowered body temperature and left blood melatonin elevated into the daylight hours (Zhdanova 2001, Journal of Clinical Endocrinology and Metabolism).
• Nine of ten jet-lag trials found melatonin taken at destination bedtime reduced jet lag after crossing five or more time zones, 0.5 mg and 5 mg worked about equally, and the number needed to treat was 2 (Herxheimer and Petrie 2002, Cochrane).
• Twenty-two of 25 melatonin gummies sold in the US missed their label by more than 10 percent, actual content ran from 74 to 347 percent of the label, and one contained no melatonin at all (Cohen 2023, JAMA).
• The American Academy of Sleep Medicine's 2017 guideline suggests clinicians not use melatonin for chronic insomnia in adults, a weak recommendation that reflects a small effect rather than a safety problem (Sateia 2017, Journal of Clinical Sleep Medicine).

Melatonin is a hormone, which puts it in a different category from almost everything else in the supplement aisle. Your body makes it on a schedule, from the same chain that makes serotonin: the amino acid tryptophan becomes serotonin, serotonin picks up an acetyl group and then a methyl group, and the result is N-acetyl-5-methoxytryptamine, the molecule on the label. The pineal gland, a pea-sized structure near the center of the brain, does most of the manufacturing, and it takes its orders from light. Darkness turns production on. Light hitting the retina, including light from a phone, turns it off (NCCIH).

That makes melatonin a signal rather than a sedative. The level in your blood starts to climb in the hours before your usual bedtime, peaks in the middle of the night, and falls toward zero by morning. Sleep researchers call the moment it begins to rise the dim light melatonin onset, or DLMO, and they use it as the most reliable marker of where a person's body clock sits (Burgess 2010, Journal of Clinical Endocrinology and Metabolism). The hormone is telling every tissue that reads it the same thing: it is night now. Whether you then sleep depends on a great deal else.

Here is the number that reframes the whole category. In 1994, a group at MIT gave 20 healthy young men oral melatonin at doses from 0.1 to 10 mg at midday and tracked their blood. The 0.1 mg and 0.3 mg doses produced peak levels inside the normal nighttime range. Every dose, including those two, shortened time to fall asleep in a controlled afternoon test and lowered oral temperature (Dollins 1994, PNAS). A tenth of a milligram is a physiologic dose. The gummies at the checkout hold 5 to 10 mg, which is 50 to 100 times that.

Older adults make less. Nighttime plasma melatonin is reduced in people over 50, which is the rationale behind most of the trials in that age group and the reason the prolonged-release prescription form used in the UK trials is indicated for people 55 and older (Zhdanova 2001; Wade 2010). Whether the decline causes poorer sleep or simply travels with it is still argued, and in the 2001 trial the healthy sleepers over 50 had low melatonin too and did not sleep any better on it.

What happens after you swallow a tablet is quick and inefficient. Only about 15 percent of an oral dose reaches the bloodstream, because the liver pulls most of it out on the first pass. What does get through peaks at roughly 50 minutes and has a half-life of about 45 minutes, so an immediate-release dose is largely cleared within three to four hours (Harpsøe 2015, European Journal of Clinical Pharmacology). The liver enzyme doing the clearing is CYP1A2 (Härtter 2001), and that is why caffeine, smoking, oral contraceptives, and the antidepressant fluvoxamine all change how long melatonin lingers (Harpsøe 2015). The short half-life is the whole argument for low doses. A 0.5 mg dose mimics the natural pulse and is gone by morning. A 10 mg dose does not bind more receptors; it keeps them occupied longer, sometimes into the next day.

Food is not where this comes from. Supplement melatonin is manufactured synthetically in nearly every case, and the few exceptions are sourced from animal or microbial material (NCCIH). "Natural" on a melatonin label describes the molecule, which is identical either way, not the source.

The regulatory split is the last thing worth knowing before the evidence. In the United States, melatonin is a dietary supplement under DSHEA, sold next to the vitamins with no dose ceiling. In several other countries it is a prescription drug (NCCIH). Same molecule, two legal systems, and most of the quality problems lower on this page trace to the first one. The habits that make any of it work, a dark room, a consistent bedtime, and less screen light after dusk, are covered in how to get the sleep your body needs.

How it works

Melatonin works by binding two receptors, MT1 and MT2, on the surface of cells in the suprachiasmatic nucleus, the cluster of neurons in the hypothalamus that keeps the body's master clock. Both are G protein-coupled receptors, meaning they pass a message into the cell rather than opening a channel, and both are high-affinity, meaning they respond to the small amounts the pineal gland releases at night (Liu 2016, Annual Review of Pharmacology and Toxicology). The same two receptors are the targets of the prescription drugs ramelteon and tasimelteon, which is a useful reminder that this is drug pharmacology sold in a supplement bottle.

The two receptors do different things, and the split explains most of what melatonin can and cannot do. In the framing of the Dubocovich group's review, MT1 signaling quiets the firing of clock neurons that keep you alert in the evening, which is the sleep-promoting effect. MT2 signaling shifts the clock's phase, moving the whole 24-hour cycle earlier or later depending on when the hormone arrives. Researchers call that second effect chronobiotic, and it is the one a jet-lagged traveler is paying for (Liu 2016).

Timing is the lever, not dose. The phase response curve is the map of what a dose does at each clock hour, and the Rush University lab built one for 0.5 mg and compared it to their earlier curve for 3 mg using the same protocol in 34 healthy adults. The low dose shifted the clock earliest when taken in the afternoon, 2 to 4 hours before the person's own melatonin onset, or roughly 9 to 11 hours before the midpoint of their sleep. Taken soon after waking, it pushed the clock later instead. When each dose was given at its own best time, 0.5 mg and 3 mg produced shifts of about the same size (Burgess 2010). Six times the dose bought nothing except a later ideal hour and a longer tail.

The sleep-promoting effect is real and small. It shows up as a drop in core body temperature and a quieter alerting signal, and in the trials it reads as minutes, not hours: about 4 minutes faster to sleep across 17 trials in one meta-analysis (Brzezinski 2005, Sleep Medicine Reviews) and about 7 minutes across 19 in another (Ferracioli-Oda 2013, PLoS One), with 8 to 13 extra minutes of total sleep. For comparison, the prolonged-release trial in adults 55 to 80 cut diary sleep latency by 24 minutes against 13 on placebo, a gap the authors described as similar to common sleep medications (Wade 2007). Melatonin is not a knockout drug. People who expect one feel nothing and reach for a bigger dose, which is how a hormone the body makes in microgram amounts ended up in 10 mg gummies.

Dose does not scale the way most people assume because the receptors are already full at a normal nighttime level. What a larger dose changes is duration. In the MIT trial of adults over 50, 0.3 mg restored sleep efficiency and brought plasma melatonin to a normal nighttime level; 3 mg also improved sleep but caused hypothermia and kept plasma melatonin elevated into the daylight hours (Zhdanova 2001). A systematic review of 16 dosing studies in adults over 55 reached the same place: higher doses stayed above threshold longer, and the authors advised the lowest possible immediate-release dose to mimic the body's own rhythm and avoid prolonged supra-physiologic levels (Vural 2014, Drugs and Aging). The Cochrane jet-lag review added one more piece: doses above 5 mg were no more effective, and a 2 mg slow-release product was relatively ineffective, which the authors read as evidence that a short-lived higher peak shifts the clock better than a long low plateau (Herxheimer and Petrie 2002).

Melatonin also has proposed roles outside the clock. The Liu review lists neuroprotection, mood, learning and memory, and antioxidant activity among the actions attributed to MT1 and MT2 signaling in laboratory and animal work. Those are proposed in the strict sense: a human trial measuring a health outcome from taking melatonin for one of those reasons would be what proves them, and none on this page does. The mechanism that is proven, a receptor signal that tells the master clock it is night, is enough to predict the outcomes below: time to fall asleep, the body clock itself, jet lag, total sleep, and the places where the signal is not the problem.

Evidence at a glance

What the evidence doesn't support yet

The sleep-aid claim is the big one and it is ahead of the data. Melatonin shortens time to fall asleep by 4 to 7 minutes on average, which is why the American Academy of Sleep Medicine's 2017 guideline suggests clinicians not use it for chronic insomnia in adults, the same weak-against rating it gave diphenhydramine and valerian (Sateia 2017). A bottle that promises a night's sleep is selling a timing signal as a sedative. The honest use is narrower: falling asleep a little sooner at a bedtime your clock resists, and moving the clock when it is in the wrong time zone.

"More is better" is the second claim, and no trial supports it. Doses above 5 mg were no more effective for jet lag (Herxheimer and Petrie 2002), 3 mg did not beat 0.3 mg on sleep efficiency in adults over 50 and left melatonin in the blood into the next day (Zhdanova 2001), and 0.5 mg shifted the clock as far as 3 mg did (Burgess 2010). The 10 mg gummy exists because it sells, not because a study asked for it.

Shift work is the third. Two 2014 reviews covering 263 and 300 shift workers found the trials small or inconclusive: perhaps 24 extra minutes of daytime sleep, no change in time to fall asleep, low-quality evidence, and no recommendation (NCCIH).

The antioxidant and anti-aging claims are the fourth. They rest on cell and animal work summarized in pharmacology reviews (Liu 2016), and no human trial on this page measured a health outcome from taking melatonin for that reason. We do not print any of the four. Timing is where the evidence is.

Clinical evidence

• Meta-analysis: melatonin for the treatment of primary sleep disorders — 19 randomized placebo-controlled trials, 1,683 adults and children with primary sleep disorders, trial lengths and doses varied. Measured sleep onset latency, total sleep time, and sleep quality. Melatonin shortened sleep latency by 7.06 minutes (95% CI 4.37 to 9.75), added 8.25 minutes of total sleep (95% CI 1.74 to 14.75), and improved sleep quality (standardized mean difference 0.22). Longer trials and higher doses showed larger latency and total-sleep effects, and the effect did not fade with continued use. Limitation: pooled across children and adults and across doses from physiologic to pharmacologic, so it cannot say which dose is right for whom.
• Effects of exogenous melatonin on sleep: a meta-analysis — 17 placebo-controlled studies, 284 people, mostly healthy adults or people with no condition other than insomnia, published 1980 to 2003. Measured sleep onset latency, sleep efficiency, and total sleep duration. Melatonin cut sleep onset latency by 4.0 minutes (95% CI 2.5 to 5.4), raised sleep efficiency by 2.2 percentage points, and added 12.8 minutes of sleep (95% CI 2.9 to 22.8). Limitation: small total sample and highly varied doses and routes across the included studies.
• Melatonin for the prevention and treatment of jet lag — Cochrane systematic review of 10 randomized trials in airline passengers, airline staff, and military personnel, oral melatonin versus placebo (one trial also versus zolpidem), doses 0.5 to 5 mg, taken at the destination bedtime of 10 PM to midnight. Measured subjective jet lag, daytime tiredness, sleep onset and quality. Nine of ten trials found melatonin reduced jet lag after flights across five or more time zones; 0.5 and 5 mg were similarly effective, people fell asleep faster after 5 mg, doses above 5 mg added nothing, and 2 mg slow-release was relatively ineffective. Number needed to treat 2. Side effects were uncommon; case reports flagged people with epilepsy and people on warfarin. Limitation: subjective outcomes and small trials, and the benefit was smaller for westward travel.
• Clinical Practice Guideline for the Pharmacologic Treatment of Chronic Insomnia in Adults: An American Academy of Sleep Medicine Clinical Practice Guideline — Clinical practice guideline built on a GRADE systematic review of randomized trials, adults with chronic insomnia, covering 14 individual agents. Measured sleep onset latency, sleep maintenance, and quality of evidence per drug. For melatonin the task force suggested clinicians not use it for sleep onset or sleep maintenance insomnia versus no treatment, a weak recommendation, placing it alongside diphenhydramine, tryptophan, and valerian; FDA-approved hypnotics received weak recommendations in favor. Limitation: a weak recommendation reflects low certainty and a small effect, not evidence of harm, and the guideline did not address jet lag or circadian timing, where melatonin's evidence is stronger.
• Optimal dosages for melatonin supplementation therapy in older adults: a systematic review of current literature — Systematic review of 16 studies, nine of them randomized controlled trials, in adults aged 55 and older (mean ages 55 to 78), oral melatonin from 0.1 mg to 50 mg/kg. Measured the effect of exogenous melatonin on blood, urine, or saliva melatonin levels. Post-dose levels rose in a dose-dependent way, older adults reached higher peaks than younger adults, and higher doses stayed above threshold longer without reaching peak any faster. The authors advised the lowest possible immediate-release dose to mimic the natural rhythm and avoid prolonged supra-physiologic levels. Limitation: pharmacokinetic outcomes, not sleep outcomes.
• Efficacy and safety of melatonin for sleep onset insomnia in children and adolescents: a meta-analysis of randomized controlled trials — Seven randomized placebo-controlled trials, 387 children and adolescents with sleep onset insomnia, short-term treatment. Measured sleep onset time, dim light melatonin onset, sleep onset latency, total sleep time, and dropouts. Sleep onset advanced 0.62 hours (about 37 minutes) more than placebo, melatonin onset advanced 0.82 hours, sleep latency fell 0.36 hours, and total sleep rose 0.38 hours; dropouts did not differ. Two serious adverse events (migraine, mild generalized epilepsy) were reported. Limitation: study quality was low to moderate, trials were short, and the authors called for more adolescent data; this is evidence for pediatrician-directed use, not for over-the-counter dosing.
• Melatonin for the management of sleep problems in children with neurodevelopmental disorders: a systematic review and meta-analysis — 13 randomized controlled trials, 682 children under 18 with neurodevelopmental disorders, melatonin versus placebo or another intervention. Measured total sleep time, sleep onset latency, night wakings, and adverse events. Total sleep time rose 48.3 minutes (95% CI 36.8 to 59.7) in nine studies and sleep onset latency fell 29 minutes (95% CI 18.2 to 39.8) in eleven; night wakings did not change; no serious medication-related adverse event was reported. Limitation: heterogeneity and inconsistency across trials kept the overall quality of evidence low, and the population is children under specialist care.
• Sleep-inducing effects of low doses of melatonin ingested in the evening — Healthy adult volunteers at MIT, double-blind placebo-controlled crossover, oral melatonin 0.3 mg or 1.0 mg or placebo at 6, 8, or 9 PM, single doses. Measured polysomnographic latency to sleep onset, to stage 2, and to REM sleep, plus next-morning mood and performance. Either dose at any of the three times shortened sleep onset latency and latency to stage 2, did not suppress or delay REM sleep, and produced no hangover on morning tests; most volunteers could tell melatonin from placebo at 6 and 8 PM. Limitation: small laboratory sample of healthy sleepers, single doses, no long-term follow-up.
• Melatonin treatment for age-related insomnia — 30 adults over 50, 15 with actigraphy-confirmed low sleep efficiency and 15 normal sleepers, double-blind placebo-controlled crossover, 0.1, 0.3, and 3.0 mg or placebo 30 minutes before bed for one week each with one-week washouts. Measured polysomnographic sleep efficiency and plasma melatonin on the last three nights of each week. The 0.3 mg dose restored sleep efficiency (P less than 0.0001), mainly in the middle third of the night, and brought plasma melatonin to a normal nighttime level; 3.0 mg and 0.1 mg also improved sleep, but 3.0 mg caused hypothermia and kept plasma melatonin elevated into the daylight hours. Normal sleepers with equally low melatonin were unaffected by any dose. Limitation: 30 people, one week per dose.
• Human phase response curves to three days of daily melatonin: 0.5 mg versus 3.0 mg — 34 healthy adults aged 18 to 42, two 5-day laboratory sessions after 7 to 9 days of fixed sleep, double-blind counterbalanced 0.5 mg melatonin or placebo once daily for 3 days at a fixed clock time, compared with the lab's earlier 3.0 mg curve from the same protocol. Measured dim light melatonin onset before and after. Maximum phase advances came when 0.5 mg was taken 2 to 4 hours before melatonin onset, or 9 to 11 hours before the midpoint of sleep; delays peaked soon after wake time; the optimal time for the lower dose was later than for 3.0 mg, and when each dose was given at its own best time the shifts were about the same size. Limitation: healthy young adults in a laboratory light schedule, three days only.
• Prolonged release melatonin in the treatment of primary insomnia: evaluation of the age cut-off for short- and long-term response — 791 adults aged 18 to 80 with primary insomnia and sleep latency over 20 minutes entered the active phase, randomized double-blind to 2 mg prolonged-release melatonin (Circadin) or placebo two hours before bed for 3 weeks, then 26 weeks with placebo completers re-randomized, then a 2-week placebo run-out; 555 completed six months. Measured diary sleep latency, Pittsburgh Sleep Quality Index, quality of life, and adverse events. At 3 weeks sleep latency fell 15.4 versus 5.5 minutes in the 55 to 80 group but the primary measure did not separate in the 18 to 80 group as a whole; improvements held or grew over 6 months with no tolerance, withdrawal, or rebound, and adverse events were mild and matched placebo. Limitation: manufacturer-sponsored, and the primary benefit was confined to adults 55 and over.
• Efficacy of prolonged release melatonin in insomnia patients aged 55-80 years: quality of sleep and next-day alertness outcomes — 354 primary-care patients aged 55 to 80 with primary insomnia in Scotland, 2-week placebo run-in then 3 weeks double-blind of 2 mg prolonged-release melatonin or placebo two hours before bed. Measured the Leeds Sleep Evaluation Questionnaire responder rate (sleep quality plus morning alertness together), Pittsburgh Sleep Quality Index, diary sleep latency, and quality of life. Responders were 26 percent on melatonin versus 15 percent on placebo (P=0.014), sleep latency fell 24.3 versus 12.9 minutes (P=0.028), and quality of life improved. Limitation: three weeks, one age group, manufacturer-sponsored, and the trial predated registration.
• Cardiovascular effects of melatonin in hypertensive patients well controlled by nifedipine: a 24-hour study — 47 adults with mild to moderate hypertension controlled on nifedipine GITS 30 or 60 mg each morning for at least 3 months, double-blind crossover, 5 mg melatonin or placebo at 10:30 PM for 4 weeks each. Measured 24-hour ambulatory blood pressure and heart rate. Evening melatonin raised systolic pressure by 6.5 mmHg (P less than 0.001), diastolic by 4.9 mmHg (P less than 0.01), and heart rate by 3.9 beats per minute, with the diastolic and heart-rate rise most evident in the morning and afternoon. The authors proposed competition between melatonin and the calcium channel blocker. Limitation: one drug, one dose, 47 people; other trials in people not on calcium channel blockers have reported lower nighttime pressure, so this is a drug-specific signal.
• Acute melatonin administration in humans impairs glucose tolerance in both the morning and evening — 21 healthy women, mean age 24, placebo-controlled single-blind crossover on four non-consecutive days: 5 mg melatonin or placebo 15 minutes before a 75 g oral glucose tolerance test at 9 AM and at 9 PM. Measured plasma glucose and insulin responses. Melatonin raised the incremental glucose area under the curve by 186 percent in the morning and 54 percent in the evening, and peak glucose by 21 and 27 percent, by reducing insulin release in the morning and insulin sensitivity in the evening. Limitation: healthy young women, one acute 5 mg dose, and no trial in people taking glucose-lowering medication.
• Increased bioavailability of oral melatonin after fluvoxamine coadministration — Five healthy men, one a CYP2D6 poor metabolizer, pharmacokinetic crossover of 5 mg oral melatonin with and without 50 mg fluvoxamine, serum sampled for 28 hours. Measured melatonin and fluvoxamine concentrations. Fluvoxamine raised the melatonin area-under-the-curve 17-fold and the peak concentration 12-fold without changing the terminal half-life, pointing to blocked first-pass clearance rather than increased production; the poor metabolizer showed a larger, longer effect. Limitation: five volunteers, one dose of each drug; the mechanism (CYP1A2 inhibition) was confirmed separately in liver microsomes.
• Pediatric Melatonin Ingestions - United States, 2012-2021 — Retrospective analysis of the National Poison Data System, all isolated melatonin ingestions reported in people aged 19 and under, January 2012 to December 2021. Measured counts, outcomes, and hospitalizations. 260,435 pediatric ingestions were reported, the annual number rose 530 percent, melatonin's share of all pediatric ingestion reports rose from 0.6 to 4.9 percent, and hospitalizations and serious outcomes increased, driven by unintentional ingestions in children aged 5 and under; five children required mechanical ventilation and two died. Melatonin became the most frequently ingested substance in pediatric poison-center reports in 2020. Limitation: poison-center data cannot confirm cause of death or dose ingested.
• Melatonin Natural Health Products and Supplements: Presence of Serotonin and Significant Variability of Melatonin Content — Laboratory analysis of 31 commercial melatonin supplements across brands and dosage forms purchased in Canada, by liquid chromatography with electrochemical and mass-spectrometric detection. Measured melatonin content against label and screened for serotonin. Melatonin ranged from 83 percent below the labeled amount to 478 percent of it, lot-to-lot content within one product varied by as much as 465 percent, more than 71 percent of products missed label by over 10 percent, and serotonin (a controlled substance) was found in 8 products at 1 to 75 micrograms. Variability did not track manufacturer or product type. Limitation: Canadian market, one time point, and the products are not named.
• Quantity of Melatonin and CBD in Melatonin Gummies Sold in the US — Laboratory analysis of 25 melatonin gummy brands listed in the NIH Dietary Supplement Label Database and purchased online in September 2022, by liquid chromatography. Measured melatonin, cannabidiol, and serotonin against label. One product contained no detectable melatonin but 31.3 mg of CBD; the rest ranged from 1.3 to 13.1 mg per serving, 74 to 347 percent of the labeled amount; 22 of 25 (88 percent) were inaccurately labeled and only 3 were within 10 percent of the claim; five declared CBD at 10.6 to 31.3 mg; no serotonin was found. The authors noted that 0.1 to 0.3 mg raises young adults to a nighttime level, so a labeled gummy serving can deliver 40 to 130 times that. Limitation: one sample per brand, gummies only, and no batch-to-batch data.

Dosage & forms

• Fast-dissolve low dose (0.5 to 1 mg dots or orally disintegrating tablets):
• Standard immediate-release tablet or capsule, 3 mg:
• Timed-release (prolonged-release) capsule or tablet, 2 to 3 mg:
• High-dose tablets and capsules, 5 to 10 mg:
• Gummies (any dose):

Safety & interactions

How much to take

The research dose for falling asleep is 0.3 to 1 mg, and the common product dose is 3 to 10 mg. Those do not match. In the MIT trials, 0.3 and 1 mg shortened time to fall asleep when given at 6, 8, or 9 PM with no hangover the next morning (Zhdanova 1995), and 0.3 mg taken 30 minutes before bed restored sleep efficiency in adults over 50 while 3 mg caused hypothermia and left blood melatonin elevated into daylight (Zhdanova 2001). The systematic review of dosing in adults over 55 landed on the lowest possible immediate-release dose (Vural 2014). Start at 0.5 to 1 mg. If it does nothing after three nights at the right time, the problem is more likely the time than the amount.

Jet lag is the exception where the studied range runs higher. The Cochrane review found 0.5 to 5 mg similarly effective, with people falling asleep faster and sleeping better after 5 mg than 0.5 mg, and nothing more from doses above 5 mg (Herxheimer and Petrie 2002). Take it at the destination's bedtime, between 10 PM and midnight, starting the first night there. Taken at the wrong time, early in the day, it causes sleepiness and slows adaptation.

For moving your clock earlier at home, the dose is smaller and the hour is earlier than most people expect. In the Rush University phase-response study, 0.5 mg produced its largest advance when taken 2 to 4 hours before the person's natural melatonin onset, which works out to roughly 9 to 11 hours before the midpoint of sleep, so late afternoon or early evening for someone who sleeps midnight to 8 (Burgess 2010). A 3 mg dose wants an earlier hour still. Expect three nights before the shift is measurable.

For waking in the night, the trial evidence is for 2 mg prolonged-release in adults 55 to 80, taken two hours before bed; in that group it held for six months with no tolerance or rebound (Wade 2010). Younger adults in the same trial did not separate from placebo on the primary measure.

What to keep it away from: alcohol, any other sedative, and a late meal. A 5 mg dose taken 15 minutes before a glucose load impaired glucose tolerance in 21 healthy women, in the morning by cutting insulin release and in the evening by lowering insulin sensitivity (Rubio-Sastre 2014, Sleep). Caffeine, smoking, and oral contraceptives change how long it lingers (Harpsøe 2015). No Tolerable Upper Intake Level has been set, and the NCCIH notes that safety at doses above what the body makes has not been fully described. The point of diminishing returns for sleep onset is somewhere between 0.3 and 1 mg. Past 5 mg nothing improves.

Safety profile

Short-term melatonin is one of the better-tolerated substances in the supplement aisle. The 2016 safety review of human studies found that short-term use was safe even at extreme doses, that the only adverse effects reported were mild (dizziness, headache, nausea, sleepiness), and that randomized trials of longer use found adverse effects comparable to placebo (Andersen 2016, Clinical Drug Investigation). The longest controlled data is the six-month prolonged-release trial: 555 adults completed it, most adverse events were mild, and there was no difference from placebo on any safety outcome, no withdrawal symptoms, and no rebound when it stopped (Wade 2010). Multi-year nightly use has not been studied in a controlled trial, in adults or children, and that gap is real.

By frequency, from the trials and the NCCIH summary: common, next-morning sleepiness or grogginess, headache, dizziness, and nausea; uncommon, lowered body temperature at 3 mg and above (Zhdanova 2001) and, in children, bedwetting, agitation, and evening urination (NCCIH); rare, published case reports of harm in people with epilepsy and in people taking warfarin (Herxheimer and Petrie 2002), and in the pediatric meta-analysis, one case each of migraine and mild generalized epilepsy among 387 children (Wei 2019).

Grogginess is a dose and timing problem, not a melatonin problem. Immediate-release melatonin has a half-life of about 45 minutes, so a 0.5 or 1 mg dose taken at 10 PM is essentially gone by 2 AM (Harpsøe 2015). At 3 mg in adults over 50, plasma melatonin was still elevated into the daylight hours (Zhdanova 2001), and older adults clear it more slowly in general (NCCIH). A 10 mg gummy at midnight is a morning-after design.

Drug interactions

Who should ask a clinician first

Pregnant and nursing women, because the safety review found no human studies and advised against use for now (Andersen 2016). Children and teenagers, every time: the trials that show benefit were run in children with specific diagnoses under medical supervision (Wei 2019; Abdelgadir 2018), the NCCIH notes open questions about effects on puberty and other hormonal development, and US poison centers logged 260,435 pediatric melatonin ingestions from 2012 to 2021, a 530 percent rise, with five children ventilated and two deaths (Lelak 2022, MMWR). Anyone on warfarin, blood pressure medication, fluvoxamine, glucose-lowering drugs, or an immunosuppressant. People with epilepsy or a seizure disorder, because of the case reports (Herxheimer and Petrie 2002; Wei 2019). Older adults with dementia, because the American Academy of Sleep Medicine's 2015 guideline recommends against melatonin in that group, and older adults generally, who clear it more slowly and feel it the next day (NCCIH). Anyone with depression taking an antidepressant, because of the fluvoxamine interaction and because sleep that does not improve on a timing signal usually has a different cause.

Signs of too much: grogginess that lasts into the morning, headache, nausea, dizziness, feeling cold. The fix is a smaller dose taken earlier, not a stronger one taken later.

If you are on any daily prescription, the question is not "is melatonin safe" but "does melatonin change what my medication does," and for blood thinners, blood pressure drugs, fluvoxamine, and glucose-lowering medication, the published answer is that it can. For ways to settle down in the evening that do not touch any of those, see ten ways to encourage your body to relax.

Buying guide

The dose: start at 0.5 to 1 mg, not 10

Match the number on the label to the trials, and the trials are low. The sleep-onset studies used 0.3 to 1 mg (Zhdanova 1995, 2001), the phase-shift study used 0.5 mg (Burgess 2010), and the jet-lag trials used 0.5 to 5 mg with nothing gained above 5 (Herxheimer and Petrie 2002). A product that lets you take 0.5 or 1 mg and add a second unit if you need it is inside the evidence. A single 10 mg tablet is not, and you cannot cut a gummy into tenths. The one place a 2 to 3 mg product earns its number is sustained release for adults over 55 who wake in the night (Wade 2010).

The form: match the release to the problem

Falling asleep and jet lag want a quick peak, so immediate-release or a fast-dissolve form taken 30 to 60 minutes before bed. Waking at 3 AM wants a lower, longer curve, so timed-release taken about two hours before bed. The Cochrane reviewers found slow-release relatively ineffective for jet lag because the clock responds to a short, higher peak (Herxheimer and Petrie 2002), so do not pack timed-release for a flight. Skip gummies regardless of dose; the reason is in the next item.

Proof of testing: the label is wrong most of the time

This is the item that matters more for melatonin than for almost any other ingredient. A Canadian lab tested 31 melatonin products and found actual content from 17 percent to 478 percent of the labeled amount, lot-to-lot swings of up to 465 percent within one product, and serotonin, a controlled substance, in 8 of them (Erland and Saxena 2017, Journal of Clinical Sleep Medicine). Six years later, a US team bought 25 melatonin gummies: 22 missed the label by more than 10 percent, actual content ran 74 to 347 percent of the claim, one contained no melatonin and 31 mg of CBD, and the "3 mg" gummy labeled "sleep plus immune" held 10.4 mg (Cohen 2023, JAMA). For a hormone that works at a tenth of a milligram, a label that is off by a factor of three is not a rounding error.

So the test for any melatonin product is a certificate of analysis for the lot in your hand, showing melatonin identity and potency against the label claim, and ideally a serotonin screen. Our melatonin is made in our own cGMP facility in Mt. Angel, Oregon; every batch is tested in our in-house lab for identity, potency, and purity; and the certificate of analysis is published by lot on our lab-testing page. Demand the same from any brand, including us. If a company cannot show you the lot number's results, the number on the front of the bottle is a guess. How to read a supplement certificate of analysis walks through what to look for.

Fillers and extras that signal a problem

Sugar, dyes, and flavors exist to make melatonin taste like candy, which is the wrong goal for a hormone that children now ingest more often than any other substance reported to poison control (Lelak 2022). "Sleep blends" that fold melatonin into a proprietary mix hide the one number that matters. CBD in a melatonin product is a drug added to a hormone, and five of the 25 gummies in the JAMA study carried it (Cohen 2023). If you want a partner for melatonin, magnesium in a chelated form, taken as its own product at its own dose, keeps both numbers visible; the magnesium chelate research page covers the form. For stress that keeps the clock from settling, the ashwagandha monograph grades the sleep-quality trials for that herb separately.

Price per day at the studied dose

Do the math per studied dose, not per bottle. Sleep Dots carry 1 mg of melatonin in a fast-dissolve dot with no sugar, dyes, or flavors, 60 dots for $15.99 on 2026-10-14 (on sale from $19.99), which is $0.27 a day at one dot or $0.53 at two. They dissolve in the mouth, so the dose is adjustable by the dot and there is no wait on digestion, and they support falling asleep at your normal bedtime and a healthy sleep-wake cycle as part of a consistent evening routine.* One dot is the physiologic dose the MIT trials used; two is the top of it.

Melatonin 3 mg Timed Release carries 3 mg in a timed-release vegetarian capsule, 60 capsules for $19.99 on 2026-10-14, or $0.33 a day. It is the one to pick for waking in the night rather than falling asleep, taken about two hours before bed, and it supports staying asleep through the night and a steady sleep-wake rhythm in adults.* It is not the one to pack for a flight. Prices and counts checked on the live product pages 2026-10-14.

Red flags

• A dose of 5 mg or more presented as the default, with no lower option.
• "Clinically proven" with no study named. The studies on this page are the ones a label can honestly point to, and they used 0.3 to 5 mg.
• A gummy, at any dose, for the reasons in the testing item.
• Melatonin inside a proprietary blend with no milligram amount.
• Timed-release sold for jet lag.
• No certificate of analysis by lot, or a certificate with no potency result against the label claim.
• Any melatonin product that does not tell blood-thinner users, pregnant women, and parents to ask first.

The bottom line

If you fall asleep later than you want to, take 0.5 to 1 mg of immediate-release melatonin 30 to 60 minutes before your target bedtime, in a dark room, for three nights before deciding whether it did anything. If you are crossing five or more time zones, 0.5 to 5 mg at the destination's bedtime, starting the first night. If you are over 55 and the problem is waking at 3 AM, 2 to 3 mg of timed-release two hours before bed has six months of trial data behind it. If you take a blood thinner, blood pressure medication, fluvoxamine, or a glucose-lowering drug, or you are pregnant, nursing, or buying for a child, the first step is a conversation, not a bottle. And if you have been taking 10 mg for a year and still cannot sleep, melatonin was never the answer to that question.

From Mt. Angel Vitamins

• Melatonin 3 mg Timed Release Sleep Support Capsules

  $19.99

  Shop Now

Frequently Asked Questions

What does melatonin actually do in the body?

Melatonin tells the body what time it is. Your pineal gland releases it when it gets dark, blood levels climb in the hours before your usual bedtime and fall to near zero by morning, and two receptors in the brain's master clock, MT1 and MT2, read that signal. MT1 quiets the clock's evening alerting, which is the mild sleep-promoting effect, and MT2 shifts the clock earlier or later depending on when the hormone arrives (Liu 2016). It is a timing cue, not a sedative. In the trials it shortens time to fall asleep by about 4 to 7 minutes on average (Brzezinski 2005; Ferracioli-Oda 2013) and moves the body clock when taken at the right hour (Burgess 2010).

Is 1 mg as effective as 3 mg or 10 mg?

For falling asleep, yes, and sometimes better. In the MIT trial of adults over 50, 0.3 mg restored sleep efficiency to normal; 3 mg also improved sleep but caused hypothermia and left melatonin in the blood into the daylight hours (Zhdanova 2001). In the phase-shift study, 0.5 mg moved the body clock as far as 3 mg when each was taken at its own best time (Burgess 2010). For jet lag, 0.5 and 5 mg were similarly effective and anything above 5 mg added nothing (Herxheimer and Petrie 2002). No trial on this page shows 10 mg doing more than 1 mg. What the larger dose changes is how long melatonin stays elevated, which is where grogginess comes from.

Why do the trials use lower doses than the products on the shelf?

Because the trials were trying to reproduce what the body does, and the products are trying to be noticed. A 0.1 to 0.3 mg dose raises blood melatonin to the normal nighttime range (Dollins 1994), so the physiology researchers at MIT tested 0.1 to 1 mg and found it worked (Zhdanova 1995, 2001). Supplement doses drifted up because melatonin is regulated as a food in the United States, with no dose ceiling, and because a hormone that works in minutes rather than hours feels like it is doing nothing to someone expecting a sleeping pill. In several other countries melatonin is prescription-only (NCCIH), and the licensed prolonged-release product in the long trials carried 2 mg (Wade 2010). The shelf dose is a marketing decision, not a research finding.

Timed release vs fast dissolve: when does each make sense?

Fast-dissolve or any immediate-release form is for falling asleep and for jet lag, because both respond to a quick, short peak; take 0.5 to 1 mg 30 to 60 minutes before bed, or 0.5 to 5 mg at destination bedtime for travel. Timed-release is for waking in the night, mostly in adults over 55: 2 mg prolonged-release two hours before bed cut sleep latency and held its gains for six months in that age group (Wade 2010), while the same form was relatively ineffective for jet lag because the clock wants a peak, not a plateau (Herxheimer and Petrie 2002). Pick by the problem: falling asleep or traveling, fast; staying asleep after 55, timed.

Does melatonin help with jet lag?

Yes, and it is the best-supported use. The Cochrane review of 10 randomized trials found that nine of them showed less jet lag when melatonin was taken at the destination's bedtime (10 PM to midnight) after flights across five or more time zones, with a number needed to treat of 2 (Herxheimer and Petrie 2002). Doses of 0.5 to 5 mg worked about equally; people fell asleep faster on 5 mg, and doses above 5 mg added nothing. Eastward flights benefit more than westward. Two cautions from the same review: taking it at the wrong time, early in the day, causes sleepiness and slows adjustment, and slow-release products worked poorly for this because the clock responds to a short peak.

Is melatonin safe to take every night?

For adults, the controlled data runs to six months: 555 adults completed a six-month trial of 2 mg prolonged-release with adverse effects matching placebo and no rebound or withdrawal when it stopped (Wade 2010), and the 2016 safety review found only mild effects (dizziness, headache, nausea, sleepiness) across short and longer studies (Andersen 2016). Multi-year nightly use has not been studied in a controlled trial, in adults or children, and the NCCIH says long-term safety information is lacking. The better question is whether you need it nightly. Melatonin is a timing signal, and if a low dose at the right hour has not fixed the problem in a few weeks, the problem is probably not timing, and a clinician is the next step rather than a bigger bottle.

Does melatonin interact with blood thinners or blood pressure medication?

It can, and both are on the ask-first list. For blood thinners, the evidence is case reports of harm in people taking warfarin with melatonin, enough that the Cochrane reviewers flagged the combination and the NCCIH says blood-thinner users need medical supervision (Herxheimer and Petrie 2002). For blood pressure, a double-blind crossover in 47 adults controlled on nifedipine found 5 mg of evening melatonin raised 24-hour blood pressure by 6.5 over 4.9 mmHg and heart rate by about 4 beats per minute (Lusardi 2000). Other trials in people not on calcium channel blockers reported lower nighttime pressure, so the direction is not predictable. Ask the prescriber before the first dose, and check home readings or INR in the first weeks.

Why does melatonin leave some people groggy?

Because the dose outlasted the night. Immediate-release melatonin has a half-life of about 45 minutes, so 0.5 or 1 mg taken at 10 PM is essentially gone by 2 AM (Harpsøe 2015). At 3 mg in adults over 50, plasma melatonin was still elevated into the daylight hours (Zhdanova 2001), and a 5 or 10 mg gummy stretches that further; older adults also clear it more slowly (NCCIH). Timing matters too: melatonin taken too late, or in the morning, is taken when the clock reads it as a delay signal and produces daytime sleepiness (Herxheimer and Petrie 2002). The fix is a smaller dose taken earlier. In the MIT trials, 0.3 and 1 mg produced no hangover on next-morning mood or performance tests (Zhdanova 1995).

Can melatonin be used for children?

Only with a pediatrician directing it. The trials that show benefit were run in children with specific diagnoses under medical supervision: sleep onset about 37 minutes earlier in children and adolescents with sleep onset insomnia (Wei 2019) and 48 more minutes of sleep in children with neurodevelopmental disorders (Abdelgadir 2018), in short trials of low-to-moderate quality. Long-term safety in children has not been studied, and because melatonin is a hormone, the NCCIH lists puberty and other hormonal development as open questions. Meanwhile US poison centers logged 260,435 pediatric melatonin ingestions from 2012 to 2021, mostly young children finding gummies (Lelak 2022), and 22 of 25 gummies tested were mislabeled (Cohen 2023). This page does not give a children's dose. A pediatrician does.

Who should ask a doctor before taking melatonin?

Anyone on warfarin or another blood thinner, blood pressure medication, fluvoxamine (which raised melatonin blood levels 17-fold in a crossover study, Härtter 2000), glucose-lowering medication, or an immunosuppressant. Anyone pregnant or nursing, because there are no human safety studies (Andersen 2016). Parents, before giving it to a child or teenager. People with epilepsy or a seizure disorder, because of case reports (Herxheimer and Petrie 2002). Older adults with dementia, for whom the American Academy of Sleep Medicine recommends against it (NCCIH). Anyone taking an antidepressant or other sedating medication. And anyone who has been using melatonin nightly for months without the sleep problem improving, because at that point the question is what else is going on, and [the evening wind-down habits](/blog/sleep-mood/supplements-feel-good-kick-stress-out) matter more than the dose.

Related Mt. Angel Vitamins Products

• Recommended for Melatonin

  Sleep Dots – Fast Acting 1 mg Melatonin Dots for Better Sleep Support

  $15.99

  Shop Now

• Recommended for Melatonin

  Melatonin 3 mg Timed Release Sleep Support Capsules

  $19.99

  Shop Now