The short answer · Tolerability
Take magnesium bisglycinate for tolerability. This is the one to buy.

Doctor's Best, High Absorption Magnesium Lysinate Glycinate, Chelated, Albion® TRAACS®, 240 Tablets (100 mg per Tablet)
4.8· 209,238 ratings· $0.16/serving
Magnesium Bisglycinate 200 mg — inside the studied 200-800 mg window
Of 857 products scored for Tolerability, Doctor's Best, High Absorption Magnesium Lysinate Glycinate, Chelated, Albion® TRAACS®, 240 Tablets (100 mg per Tablet) ranks #1 with a DoseRank score of 84. Magnesium Bisglycinate has moderate human evidence for this outcome.
The form“Why Magnesium Bisglycinate — not Ferrous Bisglycinate?”Evidence Moderate evidence
Magnesium Bisglycinate has moderate human evidence for this outcome.
- Magnesium BisglycinateThe pick
Moderate evidence · 200 mg
Chelated magnesium bound to glycine. Best-absorbed common form and gentle on digestion - the default first pick for sleep latency.
- Ferrous Bisglycinate
Strong evidence · 25 mg
Iron chelated to glycine; significantly better absorbed than sulfate with minimal constipation or nausea - the preferred form for sensitive individuals or maintenance dosing.
Questions
Frequently asked
Doctor's Best, High Absorption Magnesium Lysinate Glycinate, Chelated, Albion® TRAACS®, 240 Tablets (100 mg per Tablet) ranks #1 for Tolerability with a DoseRank score of 84. The score blends clinical strength, evidence, dose fit, trust signals and value across 857 products scored for this outcome.
Magnesium Bisglycinate has moderate human evidence for this outcome.
Magnesium Bisglycinate at 200 mg lands inside the studied 200-800 mg window. DoseRank scores this product's dose fit at 67 out of 100.
Doctor's Best, High Absorption Magnesium Lysinate Glycinate, Chelated, Albion® TRAACS®, 240 Tablets (100 mg per Tablet) works out at $0.16 per serving. $0.16 per serving — mid price tier for this outcome. The cheapest ranked pick here is $0.13 per serving.
No. Retailer links may be affiliate links, but rankings and placement are never paid. Every product is scored by the same public model, and the ranking is recomputed from the underlying product data rather than edited by hand.