Which form of magnesium is best? What the evidence actually shows
Walk into any health food shop or browse any supplement retailer and magnesium appears in more forms than almost any other mineral. Oxide, citrate, glycinate, bisglycinate, malate, taurate, threonate, chloride, carbonate, the list extends further. Each is marketed with specific claims. Glycinate is for sleep and anxiety. Threonate is for the brain. Citrate is for absorption. Oxide is cheap and inferior. Malate is for energy and muscle function. The specificity of the marketing implies a specificity of evidence that, examined closely, is often not there.
The honest picture is more nuanced. The evidence comparing magnesium forms head-to-head in clinical outcomes is limited. Most of what is known concerns bioavailability, the proportion absorbed from the gut into circulation, rather than whether different forms produce different clinical results in people who are not deficient. And bioavailability, while relevant, is not the same as clinical efficacy. What follows is an attempt to present what the evidence actually supports, form by form, with appropriate calibration of where the evidence is stronger and where it is largely assumption and marketing.
Why form matters at all
Magnesium is an element. The form of a supplement refers to what the magnesium ion is bound to, an inorganic salt in forms like oxide, chloride, and sulphate, or an organic compound in forms like citrate, glycinate, malate, and threonate. The binding compound affects how readily the magnesium dissolves in gut fluid, how efficiently it crosses the intestinal wall, whether it causes osmotic effects that accelerate gut transit, and, in some cases, whether it reaches specific tissues differently.
These are real differences. Magnesium oxide, for example, is poorly soluble at physiological pH, which limits its absorption substantially. Absorption figures in the literature vary considerably depending on dose, gastric conditions, and study design, but oxide is consistently reported as substantially lower than organic forms. The practical consequence is that a supplement with a high elemental magnesium content on the label, oxide contains around 60 percent elemental magnesium by weight, may deliver considerably less usable magnesium than a supplement with lower elemental content but higher bioavailability.
However, the clinical significance of bioavailability differences depends heavily on context. If someone is substantially deficient, getting more magnesium absorbed per dose matters more. If someone is broadly replete but seeking a specific outcome, better sleep, cognitive support, reduced muscle cramps, the question becomes whether the form studied for that outcome is the one they are taking, and the bioavailability comparison between forms becomes less important than whether the form has actual clinical trial data for the goal in question.
The organic versus inorganic distinction
A broad trend across the available evidence is that organic magnesium forms, citrate, glycinate, malate, taurate, threonate, generally show higher absorption in available studies than inorganic forms, oxide, carbonate, sulphate. A 2025 systematic review and meta-analysis found this pattern consistent across available studies, with organic forms generally showing higher bioavailability and fewer gastrointestinal adverse effects. The evidence for this trend is real but limited in terms of head-to-head trial quality and number, so it is better characterised as a consistent pattern across heterogeneous studies than a firmly established hierarchy. The likely mechanism is better solubility at gut pH and, in chelated forms like glycinate, absorption via amino acid transport pathways that bypass some of the saturable mechanisms limiting inorganic absorption.
Within organic forms, direct head-to-head comparisons are considerably less common, and the differences between them are generally smaller. Citrate and glycinate are both well absorbed; which is superior in a given person depends on individual gut chemistry and tolerance rather than a consistent hierarchy. The 2024 clinical crossover study comparing oxide, citrate, and bisglycinate directly found that both citrate and a microencapsulated oxide showed significant increases in plasma magnesium, while bisglycinate did not reach statistical significance in that particular study, a result that illustrates how individual study findings can diverge from the general literature position, and why single studies should be interpreted cautiously.
Magnesium oxide: the one worth explicitly addressing
Magnesium oxide dominates cheap, mass-market supplement products and appears in many multivitamins. It is worth being explicit about its limitations. Its low solubility means that a meaningful proportion passes through the gut without being absorbed, and its osmotic activity in the colon at higher doses produces a laxative effect that limits how much can be taken before causing diarrhoea. Across the available evidence and across dose ranges typically used in supplements, it is generally among the least efficiently absorbed of the commonly available forms, though absorption varies with dose, gastric pH, and formulation.
The main legitimate use case for magnesium oxide is as a short-term osmotic laxative, where its poor absorption is actually the mechanism of action, unabsorbed magnesium draws water into the colon. For supplementation aimed at raising systemic magnesium status, other forms are more appropriate. Products that use oxide as the primary magnesium source while prominently displaying a high elemental magnesium figure on the label are not providing as much usable magnesium as that number implies.
Magnesium citrate: the generalist
Magnesium citrate is one of the better-characterised forms in terms of bioavailability, with a Walker et al. (2003) randomised double-blind study finding it more bioavailable than oxide in direct comparison. It is well soluble, reasonably well absorbed, and widely available at moderate cost. The citrate anion contributes to the Krebs cycle and plays a role in cellular energy metabolism, though whether this produces clinically meaningful additional effects beyond the magnesium itself is not established.
Its most clinically relevant property beyond absorption is its mild laxative effect. At doses above around 300 to 400mg elemental magnesium, citrate tends to accelerate gut transit in susceptible individuals. This makes it useful for people who also want to support digestive regularity, and less suitable for those with sensitive digestion or a tendency to loose stools. It is a reasonable general-purpose choice for most people seeking to address inadequate magnesium intake.
Magnesium glycinate and bisglycinate: the tolerability argument
Magnesium glycinate, and its synonym bisglycinate, which simply specifies that two glycine molecules are bound per magnesium ion, is chelated to glycine, an amino acid. The glycine chelation allows absorption through amino acid transporters, which means it largely bypasses the osmotic effects that give oxide and citrate their laxative properties. This makes it the most widely recommended form for people who experience digestive sensitivity to other magnesium forms.
The tolerability advantage is probably its strongest clinical attribute. Whether it is absorbed more effectively than citrate in people without digestive sensitivity is not established, head-to-head comparisons show variable results, and the 2024 crossover study noted above found citrate outperforming bisglycinate in plasma magnesium increase in that particular design. There is no consistent evidence across the literature that glycinate delivers superior systemic magnesium compared to citrate in tolerant individuals, though it offers substantially better tolerability in those sensitive to the osmotic effects of other forms.
Glycinate also contributes glycine, which has its own biological activity. Glycine is an inhibitory neurotransmitter and has been studied for sleep-promoting effects at doses of around 3 grams, substantially more than the glycine content in a typical magnesium glycinate supplement, where the glycine present is closer to 1 to 1.5 grams at a 250mg elemental magnesium dose. Whether the glycine contribution in a magnesium bisglycinate supplement meaningfully adds to sleep or calming effects is plausible but not established at the doses provided. This is a case where the mechanism is biologically reasonable but the clinical evidence for a glycine-specific additive effect at supplement doses is limited.
Magnesium L-threonate: the brain bioavailability claim
Magnesium L-threonate is the most distinctively marketed and most expensive form commonly available. It was developed at MIT around 2010 specifically to raise brain magnesium concentrations, based on the observation that threonate, a metabolite of vitamin C, facilitates magnesium transport across the blood-brain barrier via glucose transporters. In animal studies, magnesium L-threonate has been shown to increase brain magnesium levels more effectively than other forms, with associated improvements in synaptic plasticity and memory performance.
The human clinical evidence has grown to include several randomised controlled trials. A 2022 trial in 109 healthy adults found improvements in multiple memory test categories after 30 days of supplementation with a magnesium L-threonate-based formula, with older participants showing greater benefit. A 2025 trial in 100 adults aged 18 to 45 found significant improvements in the NIH Total Cognition Composite and a 7.5-year reduction in estimated brain cognitive age after six weeks. A 2024 trial found improvements in self-reported sleep quality and daytime functioning at 21 days.
These are genuinely interesting findings. However, several important caveats apply. Many of the threonate trials have used multi-ingredient formulas that include phosphatidylserine, vitamins C and D, or other compounds alongside the magnesium, making it difficult to attribute effects specifically to the threonate form. Several key trials are funded by or conducted with involvement from the manufacturer (Threotech). The sample sizes are small, the 2025 trial enrolled 100 participants over six weeks, and the "7.5-year reduction in cognitive age" figure is a derived composite estimate from a short trial of this size, which should be treated with considerable caution. And the elemental magnesium content in effective threonate doses, around 108 to 144mg per day, is actually lower than what is typically provided in citrate or glycinate supplements, which means the evidence cannot straightforwardly attribute benefits to magnesium repletion per se; the threonate ligand's specific transport properties appear to be the relevant mechanism. This is a meaningfully different claim from simple bioavailability.
The cognitive evidence for threonate is preliminary. It is not dismissible, the biological rationale is plausible and there are limited but suggestive signals across a small number of trials, but neither does it constitute a robust evidence base. Large independent replication is lacking, and confident conclusions about memory improvement or cognitive protection are not supported by what currently exists.
Other forms briefly
Magnesium malate combines magnesium with malic acid, which plays a role in the Krebs cycle. It is sometimes marketed for energy and muscle function, particularly for fibromyalgia, where a small number of older trials showed some benefit. The evidence base is thin and specific to that population. Magnesium taurate combines magnesium with taurine, an amino acid with its own cardiovascular and neurological activity. Theoretical interest in taurate for cardiovascular applications exists but clinical trial evidence in humans is very limited. Magnesium chloride has good bioavailability and is used in some topical formulations, though transdermal absorption of magnesium remains poorly characterised in humans. None of these forms has a clinical evidence base comparable to citrate, glycinate, or threonate for any mainstream indication.
The bioavailability versus outcomes problem
The most important limitation in the magnesium form comparison literature is that the majority of head-to-head evidence examines bioavailability, plasma or urinary magnesium changes, rather than clinical outcomes. Outcome data comparing specific forms for specific endpoints such as sleep quality, muscle cramping, mood, cardiovascular parameters, or cognitive function is sparse across almost all form comparisons. Plasma magnesium is not a perfect surrogate for intracellular or tissue-specific magnesium, and raising plasma magnesium by a modest percentage may or may not translate into meaningful differences in the outcomes people care about.
This gap between bioavailability data and outcome data means that form recommendations based primarily on absorption studies are partly extrapolation. Baseline magnesium status is also a relevant variable: the clinical significance of absorption differences is likely greater in people who are substantially deficient than in those who are broadly replete. A form that raises plasma magnesium slightly more efficiently than another may not produce meaningfully better clinical results if both achieve adequate systemic delivery. For most people who are mildly deficient or seeking general maintenance, the practical differences between well-absorbed organic forms are probably smaller than the marketing implies.
A framework for choosing
The question of which form is best does not have a single answer independent of context. A more useful framing is to match form to purpose.
For general supplementation aimed at correcting inadequate intake, magnesium citrate is a well-studied and commonly used option with reasonable absorption and wide availability. Magnesium glycinate or bisglycinate is the more appropriate choice for anyone who finds citrate causes loose stools or digestive discomfort, and is also reasonable as a first choice for those supplementing at higher doses where tolerability becomes more relevant.
For cognitive or neurological goals specifically, magnesium L-threonate has the most form-specific evidence, though that evidence is preliminary, the trials are small and often manufacturer-funded, and the products tend to be substantially more expensive for a lower elemental magnesium dose. The signal is more credible than most supplement marketing, and more uncertain than threonate advocates typically acknowledge.
Magnesium oxide is a reasonable choice only where laxative effect is the intended outcome. For the purpose of raising systemic magnesium status it is generally among the least efficiently absorbed commonly available forms, and products relying on it as the primary source deserve scepticism regardless of the elemental magnesium figure on the label.
Whatever form is chosen, the fundamentals apply: dosing to the relevant clinical target rather than the highest number on the label, taking with food to reduce gastrointestinal effects, and recognising that in replete individuals without a specific indication the clinical benefit of supplementation itself, regardless of form, is not well established.
For detailed evidence on magnesium supplementation across outcomes including sleep, cardiovascular health, and blood sugar, see the Magnesium entry in the Evidentia library.