Each section names the population studied, because many trials enrolled older adults, poor sleepers or people with low magnesium [1][5][9]. Our evidence grading method explains what each grade requires.
Does magnesium improve sleep quality?
Evidence is mixed, with small effects at best [2][5]. A 2026 systematic review of 12 randomized trials of magnesium alone in adults found that populations, forms, elemental doses, durations and measures varied substantially [2]. Results were inconsistent across self-reported sleep, EEG, actigraphy and consumer wearables, and the reviewers rated certainty low to very low [2]. They concluded that current evidence does not support oral magnesium as a routine approach for people with insomnia; it may give modest benefits on some self-reported outcomes in some adults, but findings are inconsistent [2]. An earlier systematic review of 9 observational studies and trials (7,582 participants) found observational studies linked magnesium status with sleep quality, while the trials were contradictory [3].
A 2025 trial found a small effect. In a university-funded, placebo-controlled trial of 155 healthy adults aged 18-65 reporting poor sleep, magnesium bisglycinate (250 mg elemental magnesium a day) lowered Insomnia Severity Index scores more than placebo by week 4 (-3.9 vs -2.3; p = 0.049) [5]. The authors described the effect as small (Cohen's d = 0.2), and other psychological outcomes did not differ [5]. Exploratory analyses in that trial suggested larger improvements in people with low dietary magnesium intake [5].
Two magnesium L-threonate trials were funded by ingredient suppliers [6][7]. In a trial funded by the ingredient supplier AIDP Inc., 80 adults aged 35-55 with self-reported sleep problems took 1 g a day for 21 days, which improved Oura ring deep sleep and REM sleep scores compared with placebo [6]. Part of the difference reflected worsening in the placebo group, which declined while the magnesium group held steady [6]. AIDP commissioned the study and owns patents, three authors are paid by AIDP, and a 2025 corrigendum states the trial was registered retrospectively, on April 15, 2024 [6]. In a trial funded by the ingredient supplier Threotech Inc., 100 adults aged 18-45 with dissatisfied sleep took 2 g a day for 6 weeks, which improved self-reported sleep-related impairment, but ring-measured sleep, sleep disturbances and restorative sleep did not differ from placebo [7].
Does magnesium help people fall asleep faster?
The pooled evidence comes from older adults and is weak [1]. A 2021 meta-analysis on insomnia in older adults pooled 3 placebo-controlled trials in 151 older adults and found people fell asleep 17.36 minutes sooner after magnesium than placebo [1]. Total sleep time rose by 16.06 minutes, but that difference was not statistically significant [1]. All three trials were at moderate-to-high risk of bias and the evidence was low to very low quality [1]. The authors called the literature substandard for well-informed recommendations [1].
One included trial had 46 older adults with primary insomnia, who received 500 mg magnesium or placebo daily for 8 weeks [4]. Compared with placebo, magnesium shortened sleep onset latency and improved sleep efficiency and Insomnia Severity Index scores, though total sleep time did not differ between groups [4]. Results may not apply to healthy adults.
The newer 2026 systematic review, which included trials in people with insomnia, found results inconsistent across self-reported, EEG and actigraphy measures and rated certainty low to very low [2].
Does magnesium reduce stress?
The evidence is weak and comes from vulnerable groups [8][9]. A 2017 systematic review found 18 studies, all in samples with an existing vulnerability to anxiety: mildly anxious adults, premenstrual syndrome, the postpartum period and hypertension [8]. Four of eight studies in anxious samples reported positive effects on self-rated anxiety, but no study used a validated measure of stress, and the authors judged the evidence poor [8]. One review author was a member of a Sanofi consumer healthcare advisory board when it was published [8].
A 2018 stress trial had no placebo arm [9]. In that Sanofi-funded, single-blind trial of 264 healthy adults with high stress scores and low serum magnesium, both 300 mg magnesium a day and the same dose plus 30 mg vitamin B6 cut stress scores by 42.4% and 44.9% over 8 weeks, with no difference between the arms [9]. Without a placebo group, the trial cannot separate magnesium's effect from time and expectation [9].
Does magnesium improve memory?
Trial evidence is insufficient [10]. A 2024 NIH-funded meta-analysis of 3 randomized trials and 12 cohort studies concluded the trial evidence was insufficient to judge magnesium supplements [10]. Its cohort studies found U-shaped links between blood magnesium and cognitive impairment, which is association, not effect [10].
The supplement trials were funded by ingredient makers [7][11]. In the trial of 100 adults with dissatisfied sleep funded by the ingredient supplier Threotech Inc., 2 g a day of magnesium L-threonate for 6 weeks improved an overall NIH cognition score more than placebo (p = 0.043), with larger effects on working and episodic memory, but showed no difference from placebo on Raven's reasoning test (p = 0.953) [7]. In a trial funded by Wu-xi Magtein Bioscience, 109 healthy Chinese adults took 2 g a day of a formula combining magnesium L-threonate with phosphatidylserine and vitamins C and D, or placebo, for 30 days; the formula group improved more on all five subtests of a memory test [11]. Because that company-funded formula had several ingredients, it cannot show what magnesium alone did [11].
Does magnesium improve strength or power?
Not in athletes [14]. A 2017 meta-analysis of 14 randomized trials found no significant improvements in peak torque, muscle strength or muscle power with magnesium [14]. The authors concluded the evidence does not support a benefit for most athletes and physically active people, and that benefits appeared more pronounced in older adults and people with alcoholism [14].
What the evidence does not show
Magnesium supplements are marketed for cramps [13], but a Cochrane review of 11 trials (735 people) found that magnesium is unlikely to reduce cramps in a clinically meaningful way in older adults who have them [13]. In trials of cramps without a clear cause, largely in older adults, differences in cramp frequency between magnesium and placebo were small and not statistically significant [13]. For cramps in pregnancy the review found the trials conflicting, and it found no randomized trials of exercise-associated cramps [13].