Walk into any pharmacy and the sleep aisle is essentially a wall of melatonin. Five milligrams, ten milligrams, gummies, dissolvable strips, extended-release capsules. The implicit message is that melatonin is a sleep medicine — something you take when sleep is not happening and it makes sleep happen.
That is not what melatonin is, and it is not what it does. The gap between what most people believe about melatonin and what the research actually shows is one of the more consequential misunderstandings in popular sleep science.
Melatonin is a timing signal, not a sedative. When you use it as a sedative, it mostly does not work. When you use it as a timing signal, it can be genuinely useful — but only for a specific set of problems.
What melatonin actually does in the body
Melatonin is a hormone produced by the pineal gland in response to darkness. It does not cause sleep. It signals to the brain and body that night has begun — that it is the right time for sleep to happen. Think of it as the opening bell, not the performance itself.
The body’s melatonin production follows a precise curve: it begins rising about two hours before your habitual sleep time, peaks in the middle of the night, and drops off as dawn approaches. The rise is the signal. It tells your circadian system to lower core body temperature, suppress cortisol, and prepare for the deep-sleep stages that dominate the early part of the night.
Light suppresses it. That is the entire basis of the “screens before bed” advice — blue-spectrum light in the evening delays melatonin onset, pushing the opening bell later. The effect is real, though often overstated in popular coverage.
When melatonin actually works
The research on melatonin is consistent and fairly clear once you look at what problems it was tested for.
Jet lag. This is where melatonin has the strongest evidence. When you travel across time zones, your melatonin curve is anchored to the old time zone. Taking a small dose at the target bedtime — not at your old-timezone bedtime — helps shift the signal faster. Multiple meta-analyses support this. This is melatonin functioning as a timing signal exactly as designed.
Circadian phase delay (night owls who cannot fall asleep early). If your internal clock runs consistently late — you cannot fall asleep before 1 or 2am no matter how tired you are — a small dose of melatonin taken several hours before your target sleep time can gradually nudge the signal earlier. This is called delayed sleep phase syndrome, and melatonin has good evidence here when taken correctly.
Shift workers transitioning between schedules. Same logic: melatonin can help re-anchor the timing signal when the work schedule forces a different sleep window.
Older adults with age-related melatonin decline. Melatonin production decreases naturally with age, and older adults tend to see more benefit from supplementation than younger adults — likely because they are genuinely replacing a deficit rather than adding to an already-sufficient signal.
People in perimenopause. Some research suggests melatonin levels are disrupted during the hormonal transitions of perimenopause, making supplementation more potentially relevant here than in the general adult population. The evidence is earlier-stage than jet lag research, but the biological plausibility is strong. Perimenopause and sleep covers the fuller picture of what is shifting hormonally in that transition.
When melatonin does not work — and why
Here is the part that the supplement aisle does not tell you.
Melatonin does not work for general insomnia driven by stress, anxiety, or a racing mind. If the reason you are not sleeping is elevated cortisol, rumination, or hyperarousal, adding more of a timing signal does not fix the arousal. The door is already open. More melatonin is not the problem and more melatonin is not the solution.
Melatonin does not work for sleep debt. If you have been running an hour short every night for two weeks, a melatonin tablet will not replenish what was missed. Sleep debt is cumulative and resolves through consistent sleep, not through supplementation.
Melatonin does not reliably help new parents sleep. The problem for a parent of a newborn is almost never a misaligned circadian clock — it is fragmentation and total debt. Melatonin does not protect against wake-ups, extend sleep duration, or alter the chemistry of fragmented sleep in a meaningful way for this use case.
This is not a criticism of melatonin. It is just using the right tool for the right job. A hammer is not the wrong tool because it cannot saw wood.
The dosage problem
If you have tried melatonin and found it does not do much — or leaves you groggy the next morning — there is a good chance the dose was the problem.
Most melatonin products in the US are sold in 5mg and 10mg doses. The research on effective supplementation typically uses 0.5mg to 1mg for circadian shifting, and rarely more than 3mg even for acute jet lag use. The high-dose products are not more effective — they are just pharmacological overkill that can leave a hormone hangover and, with chronic use, may suppress the body’s own melatonin production.
Effective doses for circadian adjustment are smaller than most people assume and smaller than most products offer. If you want to experiment with melatonin for jet lag or phase shifting, look for a low-dose product (0.5mg to 1mg) and take it at your target bedtime, not at the time you are feeling tired — the timing matters more than the dose.
The timing problem
Even people who have the dose right often get the timing wrong.
Melatonin taken at bedtime when your clock is already aligned with your sleep window does almost nothing — your body is already producing it. The supplement is redundant. This is the majority of casual melatonin use: taken at 10pm because that is bedtime, when endogenous production already started at 8:30pm.
For melatonin to shift a clock, it needs to be taken before endogenous production begins, at the time you want the signal to start, not the time you want to be asleep. If you want to sleep earlier than you currently do, you take it two to three hours before your target sleep time — not when you climb into bed.
This is why many people report that melatonin “does not do anything.” It is not a failure of the supplement. It is timing that rendered it redundant.
What actually supports your own melatonin production
Your body produces melatonin reliably if you give it what it needs. The two most powerful inputs are both free:
Morning light. Bright light exposure in the first hour after waking sets the circadian anchor that drives the evening melatonin rise. The signal is: morning happened at this time, therefore evening (and melatonin onset) is this many hours later. A consistent, early morning light signal is one of the highest-leverage things you can do for your natural melatonin rhythm, and it costs nothing.
Evening darkness. Dimming lights and reducing screen brightness in the two hours before bed allows melatonin to begin rising on schedule. The effect of light on melatonin suppression is dose- and spectrum-dependent — bright overhead LED lighting delays it more than a dim lamp, blue-spectrum more than warm-toned. This does not require abandoning all screens; it requires reducing their brightness and your exposure to bright overhead lighting.
Both of these are more impactful than supplementing melatonin for the majority of people with general sleep complaints, because they work on the signal itself rather than adding an imprecise external dose.
The calm version
Melatonin is a legitimately useful tool for a specific set of problems: jet lag, night-owl circadian delay, age-related deficiency, and some perimenopausal disruption. For those problems, it works — especially at lower doses, timed correctly.
For the problems most people are trying to use it for — stress-related insomnia, fragmented-parent sleep, general “I can’t fall asleep” — it is the wrong lever. Not dangerous, just not matched to the actual mechanism. Using it there is a bit like adjusting the clock on your stove because you are late for an appointment.
The morning read in Mendtide tracks your sleep timing and consistency over time, which is the pattern that predicts whether your body’s own melatonin signal is well-anchored — and when it starts to drift. That drift is usually more informative than any supplement question.
Melatonin does not cause sleep. It announces that sleep is coming. Announce it at the right time, at the right volume, and it does exactly what it promises. Announce it too loud, too late, or for a problem it was not designed to solve, and it mostly just echoes.