Magnesium supplementation improves sleep quality

In trials, magnesium's measured effects on sleep have been small and inconsistent, and the 2026 review rated the certainty of that evidence low to very low. About a dozen randomized trials have tested magnesium on its own for sleep in adults, roughly 1,100 people in all, most of them small. A 2026 review of twelve of them found the results varied from trial to trial and from one measure to the next, rated the certainty of the evidence low to very low, and concluded that the evidence does not support magnesium as a routine treatment for insomnia, while allowing that it may give some people a modest improvement in how they rate their own sleep.

The clearest positive result comes from a 2025 trial in 155 adults who reported poor sleep: after four weeks, those taking magnesium bisglycinate rated their insomnia symptoms somewhat better than those on placebo, a small effect, with no difference on any of its other measures. Several other placebo-controlled trials found no difference, or found that sleep improved about as much on placebo as on magnesium. When sleep was measured by a device or by brain recordings rather than by asking people, there were fewer results and they did not agree. The only analysis that combined trial results, a 2021 review of three small trials in 151 older adults, found people fell asleep about 17 minutes faster than on placebo, with no clear change in total sleep time, and rated that evidence low to very low in quality.

Several of the recent trials were paid for by companies that sell the magnesium being tested. One 2024 trial of magnesium L-threonate carries a published correction. It discloses that the trial was entered into a public registry about eighteen months after it finished, so its main outcomes were never recorded before the results existed, and that the company that commissioned and funded it holds patents on the product and employs three of its six authors. In that trial, people's own ratings of their sleep improved about as much on placebo as on magnesium; the differences it reported came from a consumer sleep-tracking ring and from next-day ratings such as alertness and energy. Two authors of the 2026 review also ran two of the trials it covers, including a 2025 L-threonate trial paid for by the ingredient's supplier, and a third author who was not involved in those trials checked how they were assessed.

Whether it matters where you start is not settled. One trial's exploratory analysis found more benefit in people who reported eating less magnesium, but the 2026 review found that the trials as a whole do not show that people with lower magnesium benefit more, and few of them measured magnesium in blood or urine.

No trial has compared one form of magnesium against another for sleep, so any difference between forms comes from comparing different trials in different people. Glycinate is well absorbed and is the form in the trial with the clearest result, though that product also delivered about a gram and a half of glycine, which has its own sleep evidence, so the two cannot be separated. L-threonate has been tested in three trials, and none found a consistent improvement in how people rated their sleep; the largest, in 100 adults, improved on one of three sleep questionnaires and found no difference on a sleep-tracking ring. Oxide is poorly absorbed and was used in several of the older trials, and a trial of citrate in older adults found no difference from placebo. A product labelled only "magnesium" is usually oxide or citrate.

The way magnesium could act on sleep is well described. It dampens the brain's main excitatory signal by blocking NMDA receptors and supports its main calming signal, GABA. It also plays a role in producing melatonin and in keeping cortisol in check.

On dosing and safety, typical supplemental doses stay at or below 350mg of elemental magnesium per day. Doses above that are associated with gastrointestinal effects, and magnesium interacts with certain medications, including some antibiotics and diuretics. The trials used roughly 75mg to 730mg of elemental magnesium a day for two to twelve weeks, and larger doses did not clearly do more.