This is every compound we track, ordered as our evidence reference orders them, with four things attached to each: an evidence tier, a one-line verdict, and the population in which an effect was actually measured. Not one of them earns a strong rating for improving cognition in healthy adults — the honest top of this field is moderate, and several of the best-known names are here because they were studied properly and failed.
How to read this table
The fourth column is the one that does the work. Almost every claim you will see about these compounds is stated without a population — “improves memory,” “supports focus” — and stated that way it is false about every compound on this page. Attached to a population it can be true, and it is usually far narrower than the marketing implies: an effect found in Japanese adults aged 50 to 80 with mild cognitive impairment is not a result about a healthy 35-year-old, and an effect found in vegetarians is not a result about omnivores.
The tiers are defined in full in how we rate the evidence. In short: Strong means multiple RCTs or meta-analyses showing a consistent, clinically meaningful effect. Moderate means some RCTs with positive but limited, mixed or population-specific effects. Weak means mostly preclinical work, small studies, or inconsistent human data. None means the compound has been studied and largely failed, or has essentially not been studied in humans at all — two different situations that we distinguish in the verdict column. Avoid is reserved for ingredients that are not lawful US dietary ingredients; we will never recommend one.
Where a compound has its own page, the name links to it. Where it doesn’t, the row is the whole of what we currently publish on it, and we would rather say that than pad the catalogue.
Every compound, tiered
| Compound | Tier | Honest verdict | Where an effect was actually seen |
|---|---|---|---|
| Citicoline (CDP-choline) | Moderate | Best evidence in vascular and age-related cognitive impairment; healthy-adult data thin and often industry-funded | Older adults with age-associated memory impairment and vascular cognitive impairment. Not established in healthy high performers. |
| Alpha-GPC | Moderate Safety signal | Positive in dementia at 1,200 mg/day, mostly alongside donepezil. A large Korean cohort found ~43% higher 10-year stroke risk — see the stroke signal | Patients with mild-to-moderate Alzheimer’s or vascular dementia, especially as an add-on to donepezil. No convincing healthy-adult data. |
| Bacopa monnieri | Moderate | A real but small effect on speed of attention and rate of memory acquisition, and only after about 12 weeks. GI side effects are common | Healthy adults across a broad age range taking a standardized extract for 12 weeks or more; some signal in older adults. |
| Lion’s mane | Weak | Seven small, short trials that disagree with each other; one found worse delayed recall than placebo in young adults | Possibly Japanese adults 50–80 with mild cognitive impairment, transiently. Null to negative in healthy young adults. |
| Ginkgo biloba (EGb 761) | Moderate in dementia · None for prevention | Modest six-month benefit in dementia; little or no effect in MCI; two enormous prevention trials failed | People with diagnosed dementia, over about six months, at 240 mg/day of EGb 761. Not MCI, not healthy adults, not prevention. |
| L-theanine | Moderate (acute) | Small acute effects on reaction time and subjective calm. Not a memory agent, and not cumulative | Healthy adults, acutely, in the 0–2 hours after a dose. Some signal in anxiety-prone people and under acute stress. |
| Caffeine + L-theanine | Moderate (acute) | The best-replicated acute combination in this file. Attention-switching accuracy SMD 0.33; digit vigilance accuracy SMD 0.20, at hour 2 | Healthy adults, for a few hours. Also acutely sleep-deprived young adults in a 2025 crossover trial. |
| Omega-3 (DHA/EPA) | Moderate in subgroups · None for prevention | The classic epidemiology-versus-RCT split: excellent cohort data, trials that did not deliver in healthy older adults or established Alzheimer’s | People with mild cognitive impairment, and people with low baseline omega-3 intake. Not the general older population. |
| Creatine monohydrate | Moderate | Memory SMD around 0.29–0.31, which is small. EFSA rejected the cognitive claim | Adults 66–76; vegetarians and vegans; sleep-deprived and acutely stressed people. Not healthy young omnivores. |
| Rhodiola rosea | Weak | The fatigue trials are small and mostly low quality; cognition data are poor. Notable CYP3A4 and P-gp interaction risk | Possibly shift workers and students under exam stress. Not established for cognition itself. |
| Ashwagandha | Moderate Safety signal | A 2026 meta-analysis gives memory SMD 0.52, but the trial base is dominated by Indian studies. LiverTox likelihood score B for liver injury | Generally healthy adults 18–75, mostly in India, over about eight weeks. |
| Phosphatidylserine | Moderate in older adults · Weak for soy forms | The original bovine-derived data does not transfer cleanly to the soy-derived material now sold. FDA allowed only a heavily qualified claim | Adults over about 50 with age-associated memory impairment, particularly the poorest baseline performers. |
| Huperzine A | Moderate (China trials) | A genuine acetylcholinesterase inhibitor — pharmacologically the same class as donepezil. Cochrane calls the evidence inadequate to recommend. We discuss it; we do not recommend it | Chinese patients with Alzheimer’s or vascular dementia, in low-quality trials. Not healthy adults. |
| Vinpocetine | Avoid | FDA’s position is that it is not a lawful dietary ingredient, and it carries a 2019 FDA warning to women of childbearing age | Not applicable. We do not recommend it in any population. |
| Panax ginseng (Asian) | Weak | Cochrane found no convincing cognitive-enhancing effect in healthy people | Possibly people with subjective memory impairment or MCI, in Korean data. Not established in healthy adults. |
| American ginseng | Weak | Small acute working-memory signals, largely from one research group and partly funded by the branded extract’s maker | Healthy young adults, one to six hours after a single dose. Chronic-dosing evidence is minimal. |
| Curcumin | Weak–Moderate | One notable 18-month trial of a bioavailable formulation; meta-analyses inconsistent and entirely dependent on which formulation was used | Non-demented middle-aged and older adults, in single small trials. Not people with established Alzheimer’s. |
| Resveratrol | Weak | The cerebral blood flow changes replicate. The downstream cognitive benefit does not | Postmenopausal women, on cerebrovascular and some cognitive measures. Nobody else, reliably. |
| Acetyl-L-carnitine | Moderate (older data) | A positive 1990s meta-analysis, but the trials behind it are old, small and largely industry-run | Adults with mild cognitive impairment and mild Alzheimer’s, over three months or more. |
| Magnesium L-threonate | Weak | Almost entirely a rodent-derived rationale, plus a handful of small and mostly manufacturer-linked human trials | Older adults with cognitive impairment in one small trial; possibly people with poor sleep. Not demonstrated in healthy adults. |
| Vitamin B12 | Moderate in deficiency · None when replete | The cleanest “it depends entirely on your baseline” case here. Correcting deficiency matters enormously; supplementing replete people does nothing for cognition | People with a genuine deficiency — over-50s, vegans, people on metformin or long-term PPIs, and several clinical groups. |
| Vitamin B6 | Weak | No demonstrated cognitive benefit on its own; relevant only inside a homocysteine-lowering B-complex. Chronic high doses cause sensory peripheral neuropathy | No cognitive population on its own. The upper intake level is 100 mg/day and neuropathy has been reported below that with prolonged use. |
| Folate (B9) | Moderate in deficiency · None in general | A meta-analysis of 11 trials in ~22,000 people found no effect on cognitive aging. VITACOG found something much narrower and more interesting | Adults with MCI plus elevated homocysteine plus adequate omega-3 status. Not unselected older adults. |
| Vitamin D | Weak · None for cognition | A strong observational link and null RCTs — VitaMIND and the Finnish D Trial both found no cognitive benefit. Correcting deficiency is still justified on other grounds | Nobody, for cognition, in randomized trials. |
| Saffron | Moderate (mild-moderate AD) | Non-inferior to donepezil in small Iranian trials, with almost no independent replication outside that group | Iranian adults with mild-to-moderate Alzheimer’s; adults with mild-to-moderate depression. |
| PQQ | Weak | Two or three small Japanese trials with mixed results, and a mitochondrial-biogenesis story built on cell culture. Heavily over-marketed relative to that | Japanese adults in small, industry-affiliated trials. No large independent trial exists. |
| CoQ10 | None | Eight human cognitive trials, many null. The NIH high-dose Parkinson’s trial was terminated for futility | Not cognition. The positive signals were in progressive supranuclear palsy and chronic fatigue syndrome. |
| NAD+ precursors (NR, NMN) | Weak | The widest claim-to-evidence gap of any trending category. Blood NAD+ rises reliably; cognition does not follow | Nobody yet — multiple trials found NAD+ elevation without functional benefit, and no adequate cognitive RCT exists. |
| Taurine | Weak | Few trials, short durations, no consistent effect — and most “alertness” data come from energy-drink studies where the caffeine is doing the work | No cognitive population established. |
| L-tyrosine | Moderate (under stress only) | The clearest conditional compound here: it replenishes what depleting conditions deplete, and does nothing in their absence | Healthy young and middle-aged adults under acute stress, sleep loss, cold or multitasking load. One trial found it impaired response inhibition in healthy older adults. |
| Sage | Weak–Moderate | Genuine cholinesterase inhibition and real acute trials, but tiny ones that are rarely replicated. Concentrated S. officinalis oil contains thujone and is not a casual purchase | Healthy older volunteers, acutely; and 42 adults with mild-to-moderate Alzheimer’s in one four-month trial. |
| Gotu kola | None | A meta-analysis of 11 RCTs found no significant difference from placebo on any cognitive domain. Hepatitis case reports exist | None. Its legitimate evidence is in venous insufficiency and wound healing, not cognition. |
| Noopept (omberacetam) | Avoid | A Russian prescription drug, and not a lawful US dietary ingredient | Not applicable. We do not recommend it in any population. |
| Racetams | Avoid | Piracetam, aniracetam, oxiracetam, phenylpiracetam, pramiracetam — not lawful US dietary ingredients; FDA has issued warning letters on that basis | Not applicable. We do not recommend them in any population. |
| Caffeine (alone) | Strong for alertness · None for memory | The honest baseline every other compound here should be judged against. It works — for vigilance, not for learning | Healthy adults, for vigilance, sustained attention and reaction time. In habitual users much of the apparent benefit is reversal of overnight withdrawal. |
| Daily multivitamin-mineral | Moderate, approaching strong in 65+ | COSMOS is the only large, multi-year, replicated positive cognitive result in this entire field — and it was a supermarket multivitamin. Boring, cheap, and the most under-covered result in the niche | Older US adults over three years, with larger benefits in those with a cardiovascular disease history. |
| Cocoa flavanols | None | Tested properly inside COSMOS and failed, in the same trial where the multivitamin arm succeeded | Nobody. A possible signal survives only in the subgroup with low baseline dietary flavanol intake — hypothesis-generating, nothing more. |
| Apoaequorin (Prevagen) | None | An orally administered protein, digested to amino acids like any other. In December 2024 a federal court ordered the marketer to stop the memory claims | None. |
| Phenibut | Avoid | FDA states it is not a dietary ingredient. Dependence and a severe withdrawal syndrome are well documented — the most dangerous item commonly sold in this category | Not applicable. We do not recommend it in any population. |
| Uridine monophosphate | Weak | The synaptogenesis research behind it is genuine but mostly rodent, and it was developed clinically as part of a medical-food matrix rather than as a standalone capsule | Very early or prodromal Alzheimer’s, within the full Souvenaid matrix. No adequate standalone trial in healthy adults exists. |
What this table is not
It is not a shopping list, and the tiers are not scores to be added up. A moderate tier means the human evidence exists and is limited, not that the compound will do something for you — most of these moderates are moderate in a population you are probably not in. It is not a ranking either: caffeine sits at the top of the alertness column and is worthless for memory, which makes it better than most of this list at one job and irrelevant at another. And it is not permanent. Tiers move when trials publish, in both directions, and the date at the top of this page is when we last worked through it.
Before you use any of this to buy something, two habits are worth more than the table itself: check the label against the proprietary blend trick and the nine red-flag ingredients, and check your prescriptions against the interaction guide. Doses actually used in trials are on the individual pages, not here.
Neuro-Thrive is the only product in our audit that publishes a complete dose panel: vitamin D3 20 mcg, niacin 8 mg, vitamin B6 5 mg, bacopa monnieri 300 mg standardized to 50% bacosides, alpha-GPC 150 mg, GABA 100 mg and PQQ 10 mg. The bacopa is the actual trial dose rather than a symbolic pinch. Two things must be said before you consider it. Its PQQ is 10 mg, where the BioPQQ cognitive trial the category leans on (Nakano et al., 41 healthy elderly, 12 weeks) used 20 mg/day — that is half. And it contains alpha-GPC, which a large Korean cohort study associated with roughly 43% higher 10-year stroke risk, with a dose-response gradient; anyone on anticoagulants or with a cardiovascular or cerebrovascular history should talk to a clinician before taking it, and we would rather you read what that study found first.
We earn a commission if you buy through this link. It costs you nothing extra and it does not change what we write.
Full Neuro-Thrive review: doses, price, guarantee →
Check the current price →Related reading
- Do Nootropics Actually Work? An Honest Tier List
- How We Rate the Evidence — what each tier on this page means
- Nootropics and Antidepressants: A Compound-by-Compound Interaction Guide
- Evidence Guides — safety, interactions and how to read a label
- Myths & Debunks — the compounds on this page that failed in public
- Without Supplements — the interventions that outperform this entire table
More compound and product guides
- L-theanine: what the evidence actually shows
- Citicoline vs alpha-GPC: which has the better evidence
- Cocoa flavanols and memory: what COSMOS found
- Panax ginseng and cognition
- Curcumin and memory: bioavailability is the whole story
- CoQ10 for cognition: eight trials, half null
- Vitamin D and cognition
- Folate, homocysteine and memory
- Omega-3 dosage: what was used in the studies
- Rhodiola rosea dosage
- Saffron dosage
- Magnesium L-threonate dosage
- When to take bacopa monnieri
- When to take lion’s mane
- Huperzine A and donepezil: why not to stack them
- Multivitamin vs nootropic stack
Product reviews
- Neuro-Thrive reviews: doses, price, evidence
- NeuroPrime reviews: ingredients, price, evidence
- MemoryFuel reviews: ingredients, price, evidence
Sources
- COSMOS-Mind: multivitamin and cocoa extract effects on cognition, Alzheimer’s & Dementia
- COSMOS-Clinic and meta-analysis of the three COSMOS cognitive studies, Am J Clin Nutr
- Sagaro et al. — alpha-GPC systematic review and meta-analysis, J Alzheimers Dis
- Korean National Health Insurance cohort — alpha-GPC and subsequent stroke
- Kongkeaw et al. — meta-analysis of Bacopa monnieri RCTs, J Ethnopharmacol
- Nutrition Reviews (2025) — meta-analysis of theanine and caffeine on attention
- NIH Office of Dietary Supplements — Omega-3 fatty acids
- NIH Office of Dietary Supplements — Vitamin B12
- NIH Office of Dietary Supplements — Vitamin B6
- Clarke et al. 2014 — B vitamins and cognitive aging, ~22,000 individuals (null)
- Nutrients (2025) — Coenzyme Q10 and cognition
- QE3 high-dose CoQ10 trial in early Parkinson’s disease, JAMA Neurology (terminated for futility)
- Cognitive Vitality (ADDF) — Centella asiatica rating
- FDA — vinpocetine in dietary supplements
- FDA — phenibut in dietary supplements
- FTC statement on its win against the makers of Prevagen (December 2024)
These statements have not been evaluated by the Food and Drug Administration. Nothing here is intended to diagnose, treat, cure or prevent any disease.
This article is information, not medical advice. Talk to a doctor or pharmacist about your own situation, especially if you take prescription medication.
