Phosphatidylserine Dosage: What the Memory and Cortisol Trials Used (300-800mg)
The honest answer
300 mg/day (100 mg three times daily) is the dose used in the classic memory trials — both the bovine-cortex-PS trials (Crook 1991, Crook 1992) and the direct soy-PS replication (Jorissen 2001, no significant difference from placebo). Cortisol/exercise trials used 600–800 mg/day — well above what most single retail servings provide. There's no RDA or upper limit because PS isn't an essential nutrient.
Educational summary — not medical advice. PS has no established RDA or upper limit, and no long-term (6-month+) human safety data at any dose. This statement has not been evaluated by the FDA. This product is not intended to diagnose, treat, cure, or prevent any disease.
The dose depends on which trial's evidence you're chasing — and which ingredient
PS doesn't have one settled dose, and unlike most supplements, the bigger honesty gap here is which ingredient the dose was tested in, not just the number. Three of the trials in this evidence base used exactly 300 mg/day: Crook 1991 (bovine-cortex PS, n=149), Crook 1992 (bovine-cortex PS, n=51), and Jorissen 2001 (soy-derived PS, n=120, the direct replication of the Crook design at the same dose — and it found no significant difference from placebo on any outcome). Same number, different ingredient, different results. The cortisol/exercise-stress trials used higher doses: Starks 2008 (two of its authors are sports-supplement-industry researchers) tested 600 mg/day for 10 days in 10 healthy men (measuring the cortisol response to an exercise-stress test), and Monteleone 1992 tested 800 mg/day orally for 10 days in 9 healthy men (measuring ACTH and cortisol response to exercise). A third, older Monteleone trial (Monteleone 1990) used PS intravenously at 50-75 mg — not comparable to any oral dose, cited here only for mechanistic plausibility. The newest trial on the actual sunflower ingredient sold today (Friling 2025) tested only 100 mg/day — a third of the classic memory-trial dose — and found no significant effect on primary or secondary outcomes across the full cohort. No trial establishes that 100 mg sunflower PS is dose-equivalent to 300 mg soy or bovine PS.
| Use case | Dose studied | Result |
|---|---|---|
| Classic memory dose | 300 mg/day, 12wk | Bovine-cortex PS trials (Crook 1991/1992); a soy-PS replication at the same dose found no significant difference (Jorissen 2001) |
| Soy-PS + omega-3, combination | 300 mg PS/day (active), 100 mg maintenance, 15wk | Subgroup finding only, not the full sample (Vakhapova 2010); combination product, can't isolate PS alone |
| Cortisol/exercise stress | 600–800 mg/day, 10 days | Measured cortisol/ACTH response to exercise (Starks 2008, Monteleone 1992); no performance outcome was measured |
| Modern sunflower trial | 100 mg/day, 12wk | No significant difference on primary or secondary outcomes across the full cohort (Friling 2025, industry-funded) |
Dose checker: how much PS are you actually taking?
Enter the phosphatidylserine milligrams per serving from your label (look for "Phosphatidylserine ___ mg" on the Supplement Facts panel) and how many servings you take per day.
How to think about dose, honestly
- 300 mg/day is the dose used in the classic memory trials — but remember those trials used an ingredient (bovine-cortex PS) that supplements no longer use.
- Most 100 mg retail servings require 3 capsules/day to reach the classic cognition dose — check your label's serving size before assuming one capsule is "the trial dose."
- Retail servings are often 100-300 mg; the cortisol/exercise trials used 600-800 mg/day. Ask your clinician before choosing a dose.
- Don't treat 100 mg sunflower PS (the Friling 2025 dose) as equivalent to 300 mg soy or bovine PS. No trial has tested that equivalence, and the 100 mg sunflower trial found no significant difference on its primary outcome.
Picks by dose
Full ranking, including source and branding tiers (the primary axis for PS), is on best PS.
Discloses 300 mg per serving — one serving, no multi-capsule math needed.
Of 6 products tracked; ranked by disclosed mg, never commissions.
Frequently asked questions
How much phosphatidylserine should I take per day?
300 mg/day is the dose used in the classic memory trials. Cortisol/exercise trials used 600-800 mg/day, higher than most retail servings. No RDA or upper limit exists.
Is 300mg the "right" dose for memory?
It's the most-replicated number, but the ingredient matters more: the bovine-PS trials used 300mg (that ingredient is now discontinued), and a soy-PS trial at the same 300mg dose found no significant difference from placebo (Jorissen 2001).
Why do cortisol trials use higher doses than retail servings?
Starks 2008 and Monteleone 1992 used 600-800 mg/day, 2-2.7x the 300mg memory-trial dose and above most single retail servings.
Are there safety data at these doses?
No serious adverse events were reported in trials of 100-800 mg/day for up to 30 weeks (the longest human data). No trial has tested longer, so no long-term safety data exists.
Related
- Phosphatidylserine: what it is & who it's for
- Best PS — ranked by source/branding disclosure, then cost per mg
- PS for age-related memory
Sources
- Crook TH, Tinklenberg J, Yesavage J, et al. "Effects of phosphatidylserine in age-associated memory impairment." Neurology. 1991. PMID: 2027477 (300 mg/day, bovine-cortex PS)
- Jorissen BL, et al. "The influence of soy-derived phosphatidylserine on cognition." Nutr Neurosci. 2001. PMID: 11842880 (300/600 mg/day soy-PS, NULL)
- Starks MA, et al. "The effects of phosphatidylserine on endocrine response to moderate intensity exercise." J Int Soc Sports Nutr. 2008. PMID: 18662395 (600 mg/day)
- Monteleone P, et al. "Blunting by chronic phosphatidylserine administration of the stress-induced activation of the hypothalamo-pituitary-adrenal axis in healthy men." Eur J Clin Pharmacol. 1992. PMID: 1325348 (800 mg/day oral)
- Friling R, et al. Sunflower-derived phosphatidylserine (Sharp-PS® Green). Nutr J. 2025. PMID: 41318468 (100 mg/day, NULL primary)
- Full product dataset: /phosphatidylserine/cost-by-brand.json (CC BY 4.0).