The melatonin in most bottles is a pharmacological dose, not a physiological one. That’s the sentence I wish someone had given me before I spent two years taking 5 mg every night and wondering why I felt foggy until 10 a.m. The dose that showed up in the original sleep studies was a twentieth of what I was swallowing.
So let’s do the boring version properly: how much your body actually makes, which dose matches which problem, and when to take it – because with melatonin, timing matters more than quantity.
The dose your body makes vs the one you swallow
Blood melatonin at night peaks around 60-200 pg/mL, produced by the pineal gland in amounts measured in micrograms. A 0.3 mg capsule raises blood levels to that natural peak – and in the original randomised trial (Zhdanova et al., Nature, 1996), that dose improved sleep onset in people with insomnia. A 10 mg tablet pushes blood levels to roughly 20-30 times the physiological peak.
More is not better. It’s different: you’re using the hormone as a drug rather than as a signal.
| Dose | What it does | Where this dose comes from |
|---|---|---|
| 0.3 mg | Restores the natural nocturnal peak; improved sleep onset in the original RCT | The MIT low-dose patent (Z15-1000); the dose sleep researchers actually use |
| 0.5-1 mg | Prescription low-dose range for insomnia in older adults | Extended-release 0.75 mg, approved in the EU for adults over 55 |
| 3 mg | Standard dose in most jet-lag trials | Travel-medicine studies |
| 5 mg | Most common retail dose in the US; works, but higher than needed for most people | Shelf default |
| 10 mg | 20-30x physiological peak; no extra sleep-onset benefit, more morning grogginess and headaches | Retail “maximum strength” |
| 20 mg+ | No added benefit; side effects climb | Bulk powders and bodybuilding stacks |
One practical note on products: a 2019 analysis of 25 melatonin supplements found actual content ranging from 83% to 478% of the label claim (Cohen et al., Journal of Clinical Sleep Medicine), with gummies the least accurate. If you buy the 3 mg and get 14 mg, the “melatonin doesn’t work for me” conclusion is really a dosing accident. Look for a USP or NSF verification mark.
Timing, by goal
| Goal | Dose | When to take it |
|---|---|---|
| Sleep-onset insomnia | 0.3-1 mg | 60-90 minutes before bed (not 5 minutes before, and not at 9 p.m. if you sleep at midnight) |
| Delayed sleep phase (night owl wanting earlier) | 0.5 mg | 5-7 hours before habitual bedtime – this shifts the clock; taking it at bedtime only nudges it |
| Jet lag, eastward | 0.5-3 mg | Start 3-4 days before travel at the destination’s bedtime; continue 3-4 nights after arrival |
| Jet lag, westward | 0.5-3 mg | Take it at destination bedtime after arrival (westward is easier; morning light does more) |
| Shift work | 0.5-1 mg | 30-60 minutes before the daytime sleep, not before the shift |
| Sleep maintenance (waking at 3 a.m.) | – | Not the tool – see below |
What melatonin doesn’t do
It’s a chronobiotic, not a sedative. It tells your brain what time it is; it doesn’t knock you out. The meta-analysis numbers are honest and small: roughly 7-8 minutes faster to fall asleep and about 11 extra minutes of total sleep (Ferracioli-Oda et al., 2013), with larger effects in older adults whose own production has dropped.
Three consequences people run into:
- It won’t fix maintenance insomnia. Waking at 3 a.m. is usually alcohol, blood sugar, light, or sleep apnoea – not a melatonin shortage.
- It won’t outrun light. Bright overhead lights at 11 p.m. suppress your own melatonin more effectively than a capsule replaces it.
- It won’t compensate for a moving wake time. The single strongest sleep lever is a fixed wake time, which is why our piece on why you wake up tired even after eight hours starts with light and schedule, not supplements.
Who shouldn’t take it, and what it interacts with
Not enough evidence in pregnancy and breastfeeding, so it isn’t a default there. In children, behavioural and light strategies come first (several countries restrict it to adults under specialist guidance). It interacts with warfarin, and with CYP1A2 inhibitors such as fluvoxamine, which can multiply melatonin blood levels many times over – a 5 mg tablet in that situation behaves like a much larger one. There’s also caution in autoimmune conditions, and next-day sedation is common at high doses (the reason I stopped).
If you’re over 55 and your own production has fallen, melatonin is the supplement most likely to help. If you’re 30 and taking 10 mg because a friend’s gummy brand said so, you’re probably dosing by accident.
What I actually do
Two things, and only two. For jet lag I take 0.5 mg at the destination’s bedtime for four nights – not 5 mg, which left me reading emails at 4 a.m. the first time I tried it. The rest of the year it’s a fixed wake time, a caffeine cutoff (the numbers are in how much caffeine is too much), and dimming the flat lights an hour before bed. Training recovery is covered separately in what muscle recovery actually requires, because sleep is where the adaptation actually happens.
Evidence note: The dose and safety summary follows the US National Center for Complementary and Integrative Health (NCCIH melatonin fact sheet), with the 0.3 mg figure from Zhdanova et al. 1996 and the effect sizes from Ferracioli-Oda et al.’s 2013 meta-analysis; the product-content range is Cohen et al. 2019. Individual response varies with age, body weight and formulation, so start at the lowest effective dose rather than the shelf default.