What Melatonin Actually Is — The Biology You Need to Understand First
Melatonin is not a sleeping pill. This is the most important thing to understand before we discuss doses, and it's the reason why treating it like one (taking large doses whenever you feel like sleeping) produces poor results and unnecessary side effects.
Melatonin (N-acetyl-5-methoxytryptamine) is a hormone produced primarily by the pineal gland — a pea-sized structure deep in the brain. Its production follows a strict circadian rhythm: almost none during daylight hours, then a rapid rise beginning about 2 hours before habitual sleep time (the "dim-light melatonin onset" or DLMO), a peak around 2–4 AM, and a gradual decline before waking. This nightly secretion pattern is the biological signal your body uses to coordinate sleep timing, core body temperature reduction, cortisol suppression, and dozens of downstream physiological processes.
The word to focus on is signal, not sedative. Melatonin doesn't put you to sleep the way a sedative does — it tells your body what time it is and when to prepare for sleep. The distinction matters enormously for understanding why dose size matters so much: a signal either conveys the correct information or it overwhelms the receiver. Shouting into a phone at 120 decibels doesn't transmit a message more effectively than a normal speaking voice — it distorts it.
The Melatonin Receptor System
Melatonin exerts its effects primarily through two G-protein coupled receptors: MT1 and MT2. MT1 receptors are responsible for acute sleep promotion — they are most sensitive during the evening rise in melatonin and become partially desensitised with sustained activation. MT2 receptors are primarily responsible for circadian phase shifting — the adjustment of the biological clock's timing. Both receptor types have nanomolar binding affinity, meaning they respond to extremely low concentrations of melatonin. The physiological plasma melatonin peak of 100–200 picograms per millilitre (pg/mL) is sufficient to fully activate both receptor types. Anything above this threshold produces diminishing returns on sleep promotion while increasing receptor desensitisation, hypothermic effects, and the hormonal disruption associated with supraphysiological melatonin levels.
📊 Plasma Melatonin Levels by Dose vs Natural Production
Approximate peak plasma melatonin concentration (pg/mL) after oral supplementation. Natural nightly peak shown for reference. Data compiled from Vural et al. (2014), Arendt & Skene (2005), and Brzezinski et al. (2005).
Why Dose Is Everything — The Dose-Response Relationship
Here's what makes melatonin uniquely counter-intuitive compared to most supplements: the dose-response curve is not linear, and it actually inverts at higher doses for some outcomes. For most supplements — creatine, beta-alanine, caffeine — more produces more effect up to a saturation point. For melatonin, the story is different.
A landmark 1996 study by Dollins et al. (Journal of Clinical Endocrinology & Metabolism) administered melatonin at doses of 0.1mg, 0.3mg, 1.0mg, and 10mg to healthy young adults at noon (when endogenous melatonin is essentially zero). Both 0.3mg and 10mg reduced sleep latency by approximately the same amount — but 10mg produced significantly more daytime impairment, hypothermia, and elevated plasma concentrations that lasted well into the next day. The 0.3mg dose matched natural peak concentrations; the 10mg dose created a pharmacological level roughly 100× higher with equivalent sleep-onset benefit and substantially more side effects.
This pattern has been replicated consistently: 0.3–1mg is as effective as 5–10mg for primary sleep onset use in people with normal sleep architecture, while producing fewer next-day effects. The reason higher doses aren't more effective for sleep onset is that MT1 receptor binding is saturated at physiological melatonin concentrations. Once you've filled the receptors, adding more melatonin doesn't activate more — it just stays in circulation longer and causes hypothermia and morning grogginess.
📊 Sleep Onset Latency Reduction by Dose — Meta-Analysis Data
Minutes of sleep onset improvement vs placebo across doses. Source: Ferracioli-Oda et al. (2013) meta-analysis of 19 RCTs + updated Brzezinski et al. 2005 data. Error bars indicate 95% CI range.
0.3mg — The Physiological Dose (And the Most Under-Appreciated)
Let me state this as clearly as I can: 0.3mg is the dose with the strongest pharmacokinetic rationale for sleep onset use. It raises plasma melatonin to approximately 200–400 pg/mL — matching or slightly exceeding the natural nightly peak. It activates MT1 and MT2 receptors completely. It clears from plasma within 4–5 hours, matching the natural decline pattern. It does not produce significant morning grogginess in healthy adults.
The research pioneer on this was MIT's Richard Wurtman, who spent decades studying melatonin pharmacology and repeatedly argued — in peer-reviewed publications including NEJM correspondence and JCEM papers — that 0.3mg was physiologically appropriate and that the doses reaching consumers were pharmacologically excessive. His 2001 position paper on melatonin dosing remains one of the most cited references in the field, even though its conclusions have been largely ignored by supplement manufacturers.
Who Should Use 0.3mg
1mg — The Evidence Sweet Spot for Most Adults
If 0.3mg is the pharmacologically ideal dose, why do I call 1mg the "evidence sweet spot"? Because the RCT literature most clearly clusters around 0.5–1mg as the dose range producing the most consistent and meaningful sleep improvements across diverse populations, with a still-acceptable side effect profile. At 1mg, plasma melatonin reaches approximately 600–1,200 pg/mL — 3–6× the natural peak, but clearing plasma within 5–6 hours and producing minimal next-morning effects in most healthy adults.
The practical reason for 1mg over 0.3mg: commercially available 0.3mg products are uncommon in most markets (the US particularly). Most lowest-available-dose products are 0.5mg or 1mg. At these doses, the sleep onset benefit is clear and consistent, and the risk of the "melatonin hangover" seen with 5–10mg products is low.
Who Should Use 1mg
5mg — Moving into Therapeutic Territory
Here's where the clinical picture changes meaningfully. 5mg raises peak plasma melatonin to approximately 2,000–6,000 pg/mL — 10–30× the natural peak. For most healthy adults using melatonin for general sleep onset, this is a pharmacological overshoot. However, there are specific use cases where 5mg is the appropriate dose — and the distinction matters.
The clearest clinical indication for 5mg is severe circadian rhythm disruption: shift work disorder, non-24 sleep-wake disorder (particularly in blind individuals without light-dark cue input), and delayed sleep phase syndrome (DSPS) where the body clock is running several hours late. In these cases, the standard 0.3–1mg "chronobiotic" dose used at the wrong time relative to the misaligned circadian phase may be insufficient to produce phase shifting, and higher doses taken at a precisely timed "phase position" (rather than just "before bed") may be required.
5mg for Antioxidant and Oncological Applications
Melatonin has pleiotropic effects beyond sleep — it is a potent antioxidant, anti-inflammatory, and in research settings, has shown immunomodulatory effects at doses of 5–20mg that are being investigated in oncology. Several ongoing trials are examining melatonin at 5–20mg as an adjuvant to cancer treatment. This is entirely different from sleep promotion and operates through entirely different mechanisms (MT receptor-independent effects at high doses). These are not reasons for healthy adults to take 5mg for sleep — but they explain why 5mg products exist and have legitimate research citations.
10mg — Pharmacological Excess for General Sleep Use
I am going to be direct here: for 90% of people using melatonin to help with everyday sleep, 10mg is not just unnecessary — it is actively counterproductive. The evidence for this is not subtle or contested. It is established, consistent, and clear.
A 10mg oral dose produces peak plasma melatonin concentrations of approximately 10,000–20,000 pg/mL. Normal peak production is 100–200 pg/mL. That is a 50–100× supraphysiological level. The consequences are documented in the literature:
- Prolonged plasma half-life: 10mg oral melatonin produces elevated plasma concentrations for 12–14 hours in some individuals — meaning you are still experiencing pharmacological melatonin levels well into the following afternoon. This is the primary mechanism of "melatonin hangover."
- Receptor desensitisation: Chronic exposure to supraphysiological melatonin concentrations downregulates MT1 receptor expression in animal models. Whether this occurs in humans with long-term high-dose use is under-studied, but it raises legitimate concerns about tolerance with regular 10mg use.
- Hypothermia: Melatonin reduces core body temperature. At physiological doses, this is a normal and desirable part of sleep initiation. At 10mg, the temperature drop can be excessive and can actually disrupt sleep quality by triggering counterregulatory warming earlier than normal.
- Hormonal disruption: Supraphysiological melatonin suppresses LH and FSH pulsatility in the short term (this was explored in early contraception research). Chronic high-dose use in adolescents is a particular concern — high-dose melatonin is contraindicated in pubertal development due to its effects on reproductive hormone timing.
When Is 10mg Ever Appropriate?
There are specific, documented clinical contexts where higher melatonin doses are used under medical supervision:
- Smith-Magenis syndrome: A rare chromosomal disorder causing an inverted melatonin rhythm (melatonin produced during the day instead of night). High-dose daytime melatonin is used to suppress this abnormal pattern.
- REM sleep behaviour disorder (RBD): 3–12mg melatonin is used as a treatment, with evidence for reducing abnormal REM sleep movements.
- Oncology adjuvant therapy: Doses of 10–40mg are investigated for anti-tumour and immunomodulatory effects in cancer contexts — entirely unrelated to sleep promotion.
- Non-24 sleep-wake disorder (blind patients): Higher doses may be needed for circadian entrainment in some non-sighted individuals.
Full Dose Comparison — The Visual Summary
📊 Head-to-Head Dose Comparison — Sleep Outcomes
Relative performance across key sleep metrics by dose. Scale: 1–10 (higher = better outcome). Compiled from Ferracioli-Oda 2013, Brzezinski 2005, Dollins 1994, Zhdanova 1995, and Buscemi 2006 systematic review.
| Parameter | 0.3mg | 1mg | 3mg | 5mg | 10mg |
|---|---|---|---|---|---|
| Peak Plasma Level (pg/mL) | ~200–400 | ~600–1,200 | ~2,000–4,000 | ~3,000–6,000 | ~10,000–20,000 |
| Plasma Half-Life | 4–5 hrs | 4–6 hrs | 5–7 hrs | 6–8 hrs | 8–14 hrs |
| Sleep Onset Reduction | 6–9 min | 7–11 min | 7–12 min | 7–12 min | 7–11 min |
| Sleep Efficiency Improvement | +2–3% | +2–4% | +2–4% | +2–3% | +1–2% |
| Morning Grogginess Risk | Very Low | Low | Low–Moderate | Moderate | High |
| Hypothermia Effect | Normal (physiological) | Mild excess | Moderate | Moderate-High | Significant |
| Circadian Phase Shifting Ability | Good (timed correctly) | Good | Good–Strong | Strong | Strong but over-suppresses morning rise |
| Receptor Desensitisation Risk | Minimal | Low | Low–Moderate | Moderate | Higher with chronic use |
| Appropriate For General Sleep? | ✔ Yes — ideal | ✔ Yes — effective | ✔ With monitoring | ⚠️ Specific cases only | ✗ Not recommended |
| Evidence Quality for Sleep | Multiple RCTs | Multiple RCTs | RCT evidence | Mixed (population-specific) | Exceeds studied clinical utility |
Timing Matters as Much as Dose — The Circadian Clock Explanation
Here's the thing I see missing from almost every melatonin discussion: the dose you take matters less than when you take it relative to your internal circadian phase. This is not a minor nuance — it's probably the most practically important piece of melatonin science for everyday users.
Your circadian clock runs on approximately a 24-hour cycle, anchored by light exposure. The key moment in melatonin pharmacology is the DLMO — dim-light melatonin onset, the time at which endogenous melatonin begins its evening rise in the absence of light. This typically occurs about 2 hours before your habitual sleep time. For most people who sleep at 11pm, DLMO is around 9pm.
Melatonin supplementation has very different effects depending on when it is taken relative to DLMO:
🌙 Melatonin Timing Effect on Circadian Phase — Relative to Natural DLMO
The practical implications of this timing science:
- For falling asleep faster (sleep onset latency): Take melatonin 30–60 minutes before your desired sleep time. This is the standard recommendation and is what most RCTs measure.
- For resetting a delayed body clock (night owl syndrome, jet lag travelling east): Take melatonin 5–6 hours BEFORE your natural DLMO — potentially as early as 3–5pm. This produces maximum phase-advancing effect because you're supplementing into a circadian phase where melatonin would not normally be present.
- For jet lag travelling west: The body clock delays more easily than advances. Melatonin taken late in the evening at the destination helps accelerate the delay. Dose: 0.5–3mg at destination bedtime for 3–4 nights.
- The timing error that makes people think melatonin "doesn't work": Many people take melatonin immediately before bed — which is 30 minutes after their DLMO, meaning their body has already started producing melatonin naturally. At this point, supplemental melatonin adds relatively little. Taking it 60–90 minutes before the intended sleep time produces meaningfully better results.
📊 Sleep Onset Improvement by Timing vs Dose — 0.5mg Study
Minutes of sleep onset reduction by timing category for 0.5mg melatonin vs placebo. Early timing (2hr before bed) vs standard timing (30min before bed) vs late timing (just at bedtime). Adapted from Lewy et al. (2006) phase-response data.
Which Dose for Which Goal — The Practical Decision Framework
| Goal | Recommended Dose | Timing | Duration | Notes |
|---|---|---|---|---|
| Sleep onset (healthy adult) | 0.3–1mg | 30–60 min before bed | As needed or intermittent | Start with 0.3mg. Increase to 1mg only if no response after 1 week |
| Jet lag — eastward travel | 0.5–3mg | Destination bedtime (9–10pm local) · Start day before arrival | 3–4 nights | Eastward is harder — also use morning light at destination to anchor new time |
| Jet lag — westward travel | 0.5–1mg | Destination bedtime | 2–3 nights | Westward easier — body delays naturally · Low dose sufficient |
| Shift work (day shift after nights) | 1–3mg | 30–60 min before first day-sleep after night shift | During transition period | Block light exposure during sleep period simultaneously |
| Delayed sleep phase syndrome | 0.5–3mg | 5–6 hours before natural DLMO · NOT before bed | 3–6 weeks + lifestyle changes | Timing precision matters more than dose. Requires DLMO estimation |
| Age-related sleep fragmentation (50+) | 1–2mg slow-release | 2 hours before bed | Nightly if needed | Slow-release (PR/XR) formulation better for sleep maintenance vs onset |
| Anxiety-related sleep difficulty | 1–2mg | 60–90 min before bed | Short-term; address anxiety separately | Melatonin is not anxiolytic at these doses — address underlying anxiety |
| Antioxidant / general health | 1–5mg | Any time | Ongoing | Non-sleep use — evidence base still developing. No need for >5mg |
| Children with autism / ADHD | 0.5–3mg | 30–60 min before target sleep time | As directed by paediatric clinician | Requires physician involvement. Dose titration essential. Use lowest effective |
Real Melatonin Products — What to Actually Buy
Most melatonin products in the US are overdosed. Here is what to look for and some actual products worth considering — with an honest assessment of where they land on the dose spectrum.
Long-Term Safety, Tolerance, and Side Effects — The Honest Assessment
Short-Term Safety: Well Established
Melatonin has an excellent short-term safety profile at doses of 0.3–5mg in healthy adults. It is classified as GRAS (Generally Recognised As Safe) in the US. No serious adverse events have been reported in well-conducted trials at therapeutic doses. This good safety record is a genuine feature of the molecule — it is not a prodrug or metabolic substrate for harmful compounds.
The Grogginess Problem Is Dose-Dependent
Morning grogginess — the most common reported side effect — is almost entirely a dose and half-life problem. Studies using 0.3–1mg in appropriately timed fashion find grogginess rates indistinguishable from placebo. Studies using 5–10mg find significantly elevated next-day impairment. The solution is simple: take less, not none.
Children and Adolescents — A Special Caution
Melatonin use in children has increased dramatically — particularly for children with autism spectrum disorder and ADHD, where sleep difficulties are common. Short-term use at low doses (0.5–3mg) appears safe based on multiple paediatric RCTs. However, long-term use in children raises legitimate concerns about effects on pubertal timing, as melatonin is part of the hormonal signalling system that governs puberty onset. The 2019 systematic review (van Geijlswijk et al.) recommended caution with long-term paediatric melatonin use beyond 2–3 years and emphasised the need for physician involvement in any paediatric sleep treatment protocol.
Does Tolerance Develop?
The tolerance question for melatonin is more nuanced than for most sleep aids. Unlike benzodiazepines or z-drugs, there is no documented clinical tolerance syndrome with progressive dose escalation in healthy adults at low doses (0.3–1mg). However, at supraphysiological doses (5–10mg used chronically), animal studies show MT1 receptor downregulation. Whether this translates to clinically meaningful tolerance in humans is insufficiently studied — which is itself a reason for caution with chronic high-dose use.
Melatonin vs Sleep Medication
📊 Melatonin vs Standard Sleep Medications — Safety & Efficacy Comparison
Relative comparison of sleep onset improvement, dependency risk, and morning impairment across common sleep aids. Scale: 1–10. Sources: Multiple meta-analyses and clinical guidelines (AASM, ESS, NHS).
The 0.5mg Trial — What I Recommend Before Taking Any More
If you're currently taking 5mg or 10mg melatonin and wondering whether to make changes, here's a simple trial protocol I've used with clients: