Why Melatonin Doesn’t Work (And What Actually Does)

You take the pill. You turn off the light. You wait for the drowsy feeling that everyone promised would arrive.

Sometimes it does. Sometimes nothing much happens except a dry mouth, a vivid dream or a morning that feels heavier than it should.

That experience can make melatonin seem like a scam. Or it can make you assume that the dose was too small and reach for a stronger tablet the next night.

Neither conclusion is quite right.

Melatonin is real. Its effects are usually modest, its timing matters and its job is more specific than “switch the brain off.” It is closer to a message about biological night than to a conventional sedative. If the rest of your evening is telling your body that it is still daytime, adding a larger message does not necessarily make the system understand it better.

Melatonin is a timing signal with sleep-promoting effects

Your circadian system is coordinated by a small region in the hypothalamus called the suprachiasmatic nucleus, or SCN. It receives information about the light–dark cycle and helps time daily rhythms in alertness, temperature, hormones and sleep.

When the biological night begins, the pineal gland releases melatonin. The hormone rises in darkness, reaches a nighttime peak and declines toward morning. The signal does not contain a complete instruction to fall unconscious immediately. It tells many tissues, in effect, that the internal night has begun.

That is why sleep researchers often describe melatonin as a chronobiotic: something that communicates with, and can sometimes shift, the circadian clock. Arendt and Skene’s review describes its timing role while also acknowledging sleep-promoting effects. 1

A useful distinction is:

• Sleep pressure builds as you stay awake and is one reason you become sleepy.
• Circadian timing helps determine when your body is prepared for sleep or wakefulness.
• Melatonin is one of the signals that helps mark the biological night.

A pill may influence one part of that system without fixing the others. If you are sleeping at the wrong circadian phase, carrying a large sleep debt, exposed to bright light late at night or dealing with conditioned insomnia, melatonin may not solve the main problem.

Light can change the message

The most powerful environmental input for the circadian system is light. That does not mean every phone screen has the same effect, or that one glance at a device cancels an entire night of melatonin. Intensity, timing, spectrum, duration, distance and individual sensitivity all matter.

A classic 1995 study by Charles Czeisler and colleagues showed how the pathway can work even when ordinary vision does not. They tested six sighted people and 11 blind patients. Bright light suppressed melatonin in the sighted participants and in three blind patients whose non-image-forming light pathways were still functioning. The other blind patients did not show the same response. 2

The result was not a test of casual phone use. It was evidence that light reaching the eyes can influence the circadian pacemaker and pineal melatonin release, including in some people who cannot consciously see the light.

This is why a late-night routine can become internally contradictory. You may be trying to sleep while giving your clock a series of daytime cues: bright overhead lights, a close phone screen, work messages, a stimulating show and a changing bedtime. A melatonin supplement may add a night-related signal, but it does not make the other cues disappear.

The practical target is not darkness as a moral achievement. It is a gentler transition: dimmer light in the period before sleep, less stimulating content and a wake time that gives your circadian system a reliable anchor.

The “natural dose” comparison needs care

The body’s own nighttime melatonin production is often estimated in the microgram range. One review gives an approximate estimate of 10–80 micrograms per night, but the number varies with age, timing, physiology and the assumptions used to estimate secretion. 3

Many supplements are labelled in milligrams. A milligram is one thousand micrograms, so the units can look alarming when placed side by side. But an oral dose is not delivered to the brain in a one-to-one way. Absorption, first-pass metabolism, formulation, timing and clearance all affect the blood concentration that follows a pill.

The safe conclusion is not that a 5 mg tablet is automatically toxic or that a 0.3 mg tablet is the correct amount for every person. It is that the label size is not a measure of how much your body “needs” in the abstract, and more is not automatically more effective.

What the dose studies actually say

An older meta-analysis by Brzezinski and colleagues reviewed 17 controlled studies involving 284 participants. Across dose-response data from several studies, the sleep-promoting effect generally reached a plateau at low doses, with around 0.3 mg often producing the maximum effect or an effect that was maintained or diminished at higher doses. 4

That finding helped create the popular “less is more” message. But it should not be converted into a universal prescription. The studies were small and varied in design, participants and outcomes.

A newer 2024 systematic review and dose-response meta-analysis looked at 26 randomized controlled trials, representing 1,689 observations in people with insomnia or healthy volunteers. Its model found that melatonin’s effects on sleep-onset latency and total sleep time increased across the studied range and peaked around 4 mg per day. It also suggested that taking melatonin earlier, around three hours before the desired sleep episode, could be more effective than taking it 30 minutes before bed. 5

The two findings are not necessarily a contradiction. They come from different evidence sets and models. Together, they say something more useful than either slogan:

• dose-response is not perfectly settled;
• the best dose may depend on the person, formulation, condition and outcome;
• timing can matter as much as, or more than, simply increasing the number;
• neither study supports treating 10 or 20 mg as an automatic upgrade.

A research model that peaks near 4 mg is not a personal instruction to take 4 mg three hours before bed. If you are considering melatonin, ask a doctor or pharmacist about the reason for using it, the timing, interactions and the formulation available where you live.

A 2013 meta-analysis of 19 studies involving 1,683 participants found modest average benefits compared with placebo: sleep onset was about seven minutes faster and total sleep time about eight minutes longer, with variation between studies. 6 “Modest” does not mean “imaginary.” It means melatonin is unlikely to transform every difficult night into perfect sleep.

High dose can mean more side effects, not more sleep

Menczel Schrire and colleagues reviewed high-dose melatonin studies in adults, defining high dose as at least 10 mg. The review identified 79 studies involving 3,861 participants, but only four met the prespecified low-risk-of-bias criteria for the pooled adverse-event analysis. In that limited analysis, high-dose melatonin was associated with a 1.40 risk ratio for adverse events such as drowsiness, headache and dizziness. The review did not detect a significant increase in serious adverse events or withdrawals due to adverse events. 7

That is a reason for caution, not panic. It also shows why “40% more side effects” should not be presented as a universal outcome for every person taking 10 mg. The evidence is incomplete, and many studies did not report adverse events well.

Next-day impairment still matters. If a product leaves you groggy, dizzy or less alert, do not drive, cycle in traffic or operate machinery until you know how it affects you. A supplement’s availability does not make its effects irrelevant.

The label may not tell you the whole story

There is another reason a melatonin experiment can feel inconsistent: the product may not contain exactly what the label says.

Erland and Saxena tested 31 commercial melatonin supplements purchased in Canada. The measured content ranged from 83% below the labelled amount to 478% above it. More than 71% did not fall within 10% of their label claim, and serotonin was detected in 26% of the products tested. 8

That was one sample from one market, not proof that every supplement is inaccurate. It does show why a person may have trouble interpreting a response from an unstandardized product. The same label can represent a different exposure across brands or even across lots.

If you are comparing sleep supplements, the sleep-supplement resource discussed here is an affiliate link, so this site may receive a commission if you use it. It is not an endorsement of a particular melatonin dose, and no supplement can determine whether your problem is circadian timing, insomnia, sleep apnea, medication effects or something else. Check ingredients, local regulation, third-party testing and interactions with a pharmacist or clinician rather than assuming that “natural” means risk-free.

When melatonin may make more sense

Melatonin is more logically matched to a timing problem than to the vague goal of “sleep harder.” It may be considered in situations such as:

• jet lag, where the sleep–wake schedule is temporarily out of alignment with local time;
• delayed sleep–wake phase, where the person’s biological night consistently arrives later than the desired schedule;
• some shift-work or irregular-schedule situations, when timing and light exposure are planned carefully;
• selected sleep problems in older adults or other clinical contexts, under professional advice.

These are not interchangeable. A dose and schedule that helps shift a circadian phase may not be the right approach for ordinary bedtime insomnia. A person who falls asleep at 3AM may need a different assessment from someone who falls asleep at 10PM but wakes repeatedly with loud snoring.

The timing can also be counterintuitive. Taking a hormone right at the moment you want to be asleep is not always the same as using it to shift the clock earlier. That is why a clinician or pharmacist may ask about your natural sleep window, wake time, light exposure and the time you take it.

Melatonin is also not a substitute for evaluation. Persistent insomnia may respond better to CBT-I, while breathing symptoms may require assessment for sleep apnea and a drifting schedule may need circadian treatment. The supplement cannot tell you which problem you have.

What actually helps the clock receive a clear message

The non-pill interventions are less dramatic but often more important:

1. Anchor your wake time. A reasonably consistent morning is a stronger clock cue than repeatedly trying to force an earlier bedtime.
2. Get outdoor light early in the day. Morning light can help anchor the circadian system, especially when the day begins indoors.
3. Dim the evening environment. Lower the intensity and stimulation of light in the hours before your intended sleep window. You do not need to live in complete darkness; you do need to stop making the clock guess whether it is daytime.
4. Keep the schedule plausible. A bedtime two hours earlier than your current biological night may be a demand, not a signal. Shift gradually when appropriate.
5. Protect sleep pressure. Long late naps, extended time in bed and irregular wake times can make it harder to build a clear drive for sleep.
6. Treat the reason for wakefulness. Pain, worry, medication, alcohol, temperature, restless legs and breathing problems do not disappear because melatonin is present.

These are not guarantees, and they are not a replacement for treatment. They are ways to reduce the contradiction between the signal you are taking and the environment you are creating.

Some people also like a tactile wind-down cue. Apollo Neuro is the wearable discussed here; the link is an affiliate link and the site may receive a commission if you use it. It is an optional relaxation tool, not a melatonin replacement, not a circadian treatment and not evidence that a vibration pattern can correct a sleep disorder. If a device makes you monitor yourself more intensely, it may not be helping.

Safety matters more than the size of the bottle

Melatonin is not appropriate to self-manage in every situation. Ask a health professional before using it if you are pregnant or breastfeeding, considering it for a child or teenager, taking medicines that affect bleeding, seizures, blood pressure, blood sugar or immune function, or living with a significant medical condition. Product regulation and approved uses differ by country.

If you have taken more than intended, develop severe symptoms or are worried about a child ingesting a supplement, contact local poison-control or emergency services. Keep gummies and tablets away from children; a bottle that looks like sweets is a safety problem even when the ingredient is sold as a supplement.

Speak with a clinician if sleep problems persist, interfere with daytime functioning or involve loud snoring, gasping, unusual movements, severe restless legs, major mood changes or dangerous sleepiness. If you are too sleepy to drive safely, do not drive.

The right question is not simply, “What is the strongest melatonin?” It is, “What is keeping my sleep system out of alignment, and is melatonin the appropriate tool for that problem?”

You are not broken, and the pill is not a verdict

Melatonin can help some people. It can help more in some situations than others. It can be mistimed, overinterpreted, inaccurately labelled or used to compensate for an evening full of daytime cues.

If it has not worked for you, that does not prove that your body is defective. It may mean the dose, timing, product, expectation or underlying problem is mismatched. Increasing the number is only one possible change, and often not the most informative one.

Start by making the message clearer: a stable morning, appropriate daylight, less intense evening light and a sleep opportunity that fits your actual rhythm. If you are considering melatonin, use professional advice rather than turning the bottle into a nightly guessing game.

Melatonin is a message about night. It is not a remote control for sleep.