Does magnesium help you sleep?

Walk into any Boots and you will find a shelf of magnesium marketed at people who cannot sleep. Bottles in navy and midnight blue. Moons on the label. Words like calm, unwind and night-time.

Now look closely at what those bottles actually claim. Not one of them says magnesium helps you sleep. They say things like "contributes to normal psychological function" and "reduction of tiredness and fatigue."

That is not shyness. It is the law. And once you understand why, you understand almost everything worth knowing about magnesium and sleep.

The short answer

If you are genuinely short of magnesium, topping it up may well improve your sleep. If you already have enough, there is almost no evidence that taking more does anything at all.

That distinction runs through this entire subject, and the supplement industry has every reason to blur it.

What the research actually consists of

Here is the thing that surprises people. There is no large body of magnesium sleep research. There are four randomised controlled trials. Four.

The only meta-analysis ever published on the question pooled three of them — 151 people in total.¹ That is not a research field. That is a rounding error.

The famous one

Almost every article you will read about magnesium and sleep traces back to a single 2012 Iranian study.² Forty-six people aged 60 to 75, given 500mg of elemental magnesium as magnesium oxide for eight weeks. It found improvements in insomnia severity, sleep efficiency and how long people took to fall asleep.

Now the detail that is almost always left out.

To get into that trial you had to be magnesium insufficient. The inclusion criteria required a dietary intake below 75% of the recommended amount and a serum magnesium below 0.95 mmol/L. Everybody in it was short of magnesium before they started.

So it does not show that magnesium is a sleep aid. It shows that fixing a shortfall in older adults who had one improved their sleep. Those are different claims, and only one of them is being sold to you.

Two further things worth knowing about it. Total sleep time did not improve significantly. And the participants' actual blood magnesium barely shifted — the change fell short of significance, which means the trial cannot quite demonstrate that the supplement did what it was supposed to do.

The best one is new, and modest

The strongest trial to date came out in 2025: 155 healthy adults with poor sleep, given 250mg of magnesium bisglycinate for four weeks.³

The result was positive. Just barely. Insomnia scores improved by 3.9 points against 2.3 on placebo — a difference of 1.6 points, with a p-value of 0.049. For context, 0.05 is the conventional threshold. This scraped in.

The effect size was 0.2, which is small by any standard. And every single secondary outcome was null: no significant change in fatigue, daytime sleepiness, mood, perceived stress or a second insomnia scale.

Two complications the headlines skipped. The capsules also delivered 1,523mg of glycine — an amino acid with its own sleep research behind it — so this is not a clean test of magnesium at all. And the people who benefited most were, again, the ones with the lowest magnesium intake to begin with.

And the other two

A 2010 trial of 100 older adults on magnesium citrate found sleep quality improved substantially — in both groups equally.⁴ Which is a null result for magnesium and a rather good demonstration of the placebo effect. The fourth, from 2002, involved around twelve people and was judged at high risk of bias.¹

What the reviewers made of it

The 2021 meta-analysis found magnesium shortened time-to-sleep by about 17 minutes.¹ That number gets quoted everywhere. What gets quoted rather less is the rest of the paper.

Total sleep time showed no significant benefit. Every trial was rated at moderate-to-high risk of bias. The certainty of the evidence was graded "low to very low". And the authors' own conclusion was blunt:

This review confirms that the quality of literature is substandard for physicians to make well-informed recommendations on usage of oral magnesium for older adults with insomnia.

Mah & Pitre, BMC Complementary Medicine and Therapies, 2021

A 2023 review of nine studies covering 7,582 people put the split plainly: the observational data show an association between magnesium status and sleep quality, while the randomised trials show "an uncertain association".⁵

Sleepstation, the NHS-commissioned sleep service, puts it plainly: the evidence is "neutral or slightly positive but large studies with lots of participants are needed for confirmation", and "whether taking a supplement of it will help you sleep remains to be seen."⁶

And here is a small fact that tells you a great deal. Magnesium is not mentioned on the NHS insomnia page at all.⁸ The NHS lists valerian, lavender and antihistamines among things that "cannot cure insomnia but may help you sleep better for 1 to 2 weeks". Magnesium does not make even that list.

Why the observational studies look so much better

You will find plenty of research linking higher magnesium intake to better sleep. It is real, and it is worth almost nothing as evidence that supplements work.

High-magnesium diets are diets full of wholegrains, leafy greens, legumes and nuts. People who eat that way differ from people who do not in a hundred other respects — weight, exercise, alcohol, income, whether they eat at 11pm. Untangling magnesium from all of that is not possible in an observational study.

There is also a genuine chicken-and-egg problem that a 2025 mechanistic review flagged: poor sleep may lower magnesium rather than the other way round.⁹ Nobody has established which direction the arrow points.

So are you actually short of it?

This is the question worth asking, and the honest answer is that it is harder to pin down than the supplement aisle suggests. The National Diet and Nutrition Survey measures how many people fall below the lower reference nutrient intake — the level below which intake is almost certainly inadequate — and it has reported low magnesium intakes in Britain for years. The survey’s own summary of its first nine years records “evidence of low intakes of magnesium for children aged 11 to 18 years, adults aged 19 to 64 years and adults aged 65 years and over”.¹⁰

We are not going to print a single percentage here, and it is worth explaining why. We tried to pin the figures down and could not reconcile them. The published NDNS reporting rounds do not agree with each other: one round puts men aged 19 to 64 below the LRNI at around one in fifty, another at roughly one in eight; one puts older adults in low single figures, another has more than a quarter of women aged 75 and over below it.^1010b Some of that spread is probably real — survey rounds sample different people and the food composition tables behind them change — but we could not verify which figures apply to which round without the underlying data tables, so we are not going to quote one as though it settled the matter. Anyone citing a confident single number for British magnesium deficiency should be asked which round it came from.

Two things do survive that mess, and they are the ones that matter.

Teenage girls come out badly in every round we looked at — consistently around half below the LRNI, by some distance the worst-served group in the survey. That finding is stable enough to take seriously.

And the survey itself cautions against reading too much into any of it. NDNS notes that the health implications of intakes below the LRNI for magnesium are unclear, because the UK dietary reference values for magnesium rest on limited data in the first place.^10b That is the government’s own nutrition survey saying the yardstick is soft. It is a useful thing to hold in mind when a label implies you are running a deficit against a firm number.

There is a further reason not to read a population statistic as a statement about yourself. The one trial that produced a clean positive result recruited people who had been screened for insufficiency — it was a repletion study, and repleting someone who is short of a nutrient is a different proposition from topping up someone who is not.

A blood test is less help than you would hope. Under 1% of the body's magnesium is in the bloodstream, so a normal serum reading does not prove your stores are adequate.⁷ There is no cheap, reliable way to know where you stand.

How much does the form matter?

Glycinate. Bisglycinate. Citrate. L-threonate. Malate. Oxide. The premium end of the market is built on the idea that the form matters enormously, and specifically that some forms are better for sleep.

We could not find a head-to-head trial of two magnesium forms with sleep as an outcome. If one has been published we would genuinely like to see it. On the evidence we could locate, every claim that glycinate beats citrate for sleep is extrapolation from absorption data rather than a demonstrated difference in how you sleep — which is not the same as saying the difference is not there, only that nobody appears to have measured it.

Note which forms the actual sleep trials used. Oxide and citrate — the two that premium brands now dismiss as inferior.

Absorption differences are real but smaller than advertised. A crossover study found magnesium citrate outperformed oxide on urinary excretion and short-term serum levels — but found no difference in intracellular magnesium, which is arguably what matters.¹¹ The study was funded by a magnesium manufacturer.

What better absorption genuinely buys you is comfort. Poorly absorbed magnesium stays in the gut and acts as a laxative. So "gentle on the stomach" is a fair claim for glycinate. "Better for sleep" is not.

Magnesium L-threonate deserves a particular note. It is sold on its ability to raise magnesium levels in the brain — evidence that comes from rodent studies, not humans. Its elemental magnesium content is very low, roughly 75mg per gram, so a heavily marketed "1,000mg" dose delivers less magnesium than a cheap tablet. And the main human sleep trial of it states that it was funded by the supplier of the branded ingredient while simultaneously declaring that the authors have no competing financial interests.¹² That contradiction is printed on the face of the paper.

Sprays, flakes and Epsom salts

Here the answer is not "uncertain". It is no.

A 2017 review in Nutrients examined the transdermal magnesium literature and concluded that "the propagation of transdermal magnesium is scientifically unsupported".¹³ The supporting studies were tiny — some as small as nine people — and several measured magnesium in hair rather than blood. The most-cited positive result was a pilot study of 25 people that reached significance only within a subgroup.¹⁴

An Epsom salt bath may well be relaxing. A hot bath before bed has its own modest evidence behind it. But the magnesium is not getting through your skin in any meaningful quantity.

How much, and who should be careful

The NHS sets the reference intake at 300mg a day for men aged 19–64 and 270mg for women, and advises that "having 400mg or less a day of magnesium from supplements is unlikely to cause any harm."¹⁵ Above that, the usual consequence is diarrhoea. The UK Expert Group on Vitamins and Minerals set a guidance level of 400mg for supplemental magnesium on exactly that basis.¹⁶

All of these limits apply to supplements only. Magnesium from food carries no upper limit — healthy kidneys excrete the excess.

Why nobody in Britain is allowed to tell you it works

This is the part almost no article covers, and it is the most useful thing on this page.

In Great Britain, a food or supplement may only carry health claims that appear on the official Nutrition and Health Claims Register.¹⁷ Magnesium has ten authorised claims. They cover tiredness and fatigue, electrolyte balance, energy metabolism, nervous system function, muscle function, protein synthesis, psychological function, bones, teeth and cell division.

Sleep is not among them. There is no authorised claim in Britain that magnesium does anything for sleep, because the evidence submitted has never been strong enough to get one approved.

Which is why the labels talk about "psychological function" and "tiredness" next to a picture of the moon. The claim is legal. The implication is what you are meant to take away.

The Advertising Standards Authority has ruled on this repeatedly. In July 2024 it upheld a complaint against a magnesium "Night-Time Gummies" product, finding there were "no authorised health claims relating to sleep quality on the GB Register", so phrases like "prepare for a good night's sleep" and "deeper sleep" were not permitted.¹⁸ It also found that dropping the words "contributes to" from the fatigue claim made it absolute, and therefore exaggerated.

A year later it upheld against a magnesium-containing product advertised as "the best sleep supplement on the market."¹⁹ In March 2026 it upheld again, restating that only claims on the GB register are permitted.²⁰

What the evidence does support

If you are not sleeping, the intervention with by far the strongest evidence behind it is not a supplement. It is cognitive behavioural therapy for insomnia. NICE puts CBT-I first in the treatment pathway for long-term insomnia, ahead of any medication.²¹ It is available on the NHS in parts of the country, and through self-help programmes.

That is a less appealing sentence than "take this capsule." It is also the honest one.

The bottom line

Magnesium is essential, plenty of people in Britain do not get enough of it, and correcting a genuine shortfall may improve how you sleep. That much is fair.

Beyond that, the evidence thins out fast. Four small trials. One meta-analysis rating the certainty as low to very low and calling the literature substandard. The best trial scraping significance with every secondary measure null. No comparative evidence for any form. No absorption at all through your skin. And no authorised claim in this country that it does anything for sleep whatsoever.

If you want to try it, a cheap magnesium citrate tablet at 200–300mg is unlikely to hurt you and costs a few pounds. Just know that you are running an experiment on yourself, not following the science. And if the sleep problem persists, the capsule is not the answer — a GP is.