Huperzine A vs Huperzia Serrata: What You’re Actually Buying

Two bottles sit side by side. One says “Huperzine A, 200 mcg.” The other says “Huperzia serrata extract, 200 mcg.” They look interchangeable. They are not. If that second extract is 1% huperzine A, a common grade, the capsule holds 2 mcg of the compound, not 200.

The confusion is understandable. Huperzine A is the active compound. Huperzia serrata is the club moss it is extracted from. So the plant name tells you where the ingredient started and nothing about how much of it made it into the capsule. The huperzine A number tells you that, and in a sample of 308 labels from the NIH Dietary Supplement Label Database, 28% listed Huperzia with no amount at all.²⁰

This guide shows how to read the label in about ten seconds, what the research says about each form, what human studies do and do not show, and what to check before buying.

Huperzine A vs Huperzia serrata at a glance

What is huperzine A?

Huperzine A (HupA) is an alkaloid, a nitrogen-containing compound made by plants. It blocks acetylcholinesterase, the enzyme that breaks down acetylcholine, a chemical messenger the brain and nerves use. With the enzyme blocked, more acetylcholine stays available between nerve cells. HupA binds the enzyme reversibly and crosses the blood-brain barrier, which is why researchers take it seriously.

It is also selective. In human blood samples, HupA blocked acetylcholinesterase at about 47 nM, while the related enzyme butyrylcholinesterase barely responded (more than 10,000 nM).²¹ That gap of more than 200-fold explains the military interest. Butyrylcholinesterase is the blood’s scavenger for nerve agents, and HupA leaves it free. Pyridostigmine, the standard pretreatment, cannot enter the brain. HupA can.

China approved HupA as an Alzheimer’s drug in the mid-1990s. In the United States it is sold as a supplement, and Operation Supplement Safety (OPSS), a Department of Defense education program, calls its legal status unclear.¹⁵

What is Huperzia serrata?

Huperzia serrata is a club moss, a low-growing, spore-bearing plant. Other names include Chinese club moss, toothed clubmoss, and firmoss. Traditional Chinese medicine uses it as Qian Ceng Ta, mainly for bruises and swelling. It grows on the ground in humid mountain forest across Asia, grows slowly, and China lists Huperzia and Phlegmariurus species as Class II protected plants.

Three name problems cause real confusion. Boswellia serrata is an unrelated resin tree. Huperzia selago, fir clubmoss, looks similar and also contains HupA. And a 2015 study treats southern Chinese plants sold as H. serrata as H. javanica, so the Latin name on a label does not guarantee a single species.

How does a moss become a supplement?

Raw moss holds very little HupA. Ma and colleagues measured it by HPLC across species, plant parts, seasons, and regions.¹ Their overall average, as reported in a 2024 Dutch RIVM review of the paper, was about 80 µg per gram of dry weight, or 0.008%.¹⁹ Four patterns stood out:

• Habitat: plants from humid forest held significantly more than plants from drier habitats.
• Season: content peaked in mid-fall (about 100 µg/g in October) and bottomed in early spring (64 to 70 µg/g in late winter).
• Plant part: leaves held 117 µg/g, stems 78, sporangia 38, and roots 25.
• Region: Yunnan plants reached 133 µg/g, while Guangdong plants held 46.

Published averages for the species range from 0.0047% to 0.025%, and two papers from one Australian group report 0.18 and 0.08 mg/g. At 80 µg/g, 200 mcg of HupA needs about 2.5 g of dried herb. A 1% extract is 40 to 200 times more concentrated than the raw moss, depending on which average is used. So a 1% label means the moss was concentrated industrially.

Reaching 1% takes concentration and purification. A Chinese industry patent (CN104262251B) describes a route to 99% pure HupA at a yield of 0.026% or more, and claims raw herb holds 0.03% to 0.06%, two to seven times the academic figures. No public source documents how commercial 1% grades are made. Whether a given powder comes purely from the plant, is boosted with isolated HupA, or both, cannot be read from its label.

Extract vs whole herb vs isolated huperzine A

Three kinds of product share the shelf, and the label rarely says which one is in the bottle.

Whole herb carries a cost beyond unpredictable dosing. In one propagation study, a plant took 15 to 20 years to grow from spore to maturity (a low-tier source, so treat the figure cautiously). Buying whole herb means buying whole wild plants. A standardized product that names its source eases that pressure.

What does the percent on the label mean?

One formula decodes any extract label. Multiply the milligrams of extract by the percentage of HupA, then by 10, to get micrograms.

A ratio and a percentage are different claims. A 30:1 ratio means 30 g of herb went into 1 g of extract. At 80 µg/g, that predicts about 2.4 mg of HupA per gram of extract, or 0.24%. Only the percentage fixes the dose, and only a lab test confirms the percentage.

Natural vs synthetic huperzine A

Natural HupA is the (-) form. Chemical synthesis often produces a racemic mix, half (-) and half (+). In rat brain tissue, the (+) form inhibits acetylcholinesterase 38 times less potently (Ki 300 nM against 8 nM), and the racemic mix is about 2 times weaker than pure (-).² A racemic capsule therefore gives roughly half the activity of the same weight of (-)-HupA.

The (+) form is not toxic. Army researchers call it non-toxic because it barely inhibits the enzyme, and tested it at 20 to 40 mg/kg in guinea pigs against the nerve agent soman.³ The real risk with a racemic product is paying for half-strength material. And “synthetic” does not always mean racemic: enantioselective routes make pure (-)-HupA, chemically identical to the plant’s. Only carbon-14 testing can tell them apart.

Testing data is thin. A 2020 industry case study found petrochemical carbon in 3 supplements that passed HPLC purity above 98%.¹⁷ It covered 3 brands, published no per-sample numbers, and has a co-author at a HupA manufacturer. It proves a detection method exists, not how common the problem is. A good certificate of analysis states (-)-HupA by HPLC and, for plant-origin claims, a carbon-14 result.

Does huperzine A work? What the human data shows

The best evidence is in Alzheimer’s disease, and it is weak. A 2008 Cochrane review pooled 6 trials with 454 patients and found inadequate evidence to recommend huperzine A.¹⁰ A 2013 meta-analysis counted 20 trials with 1,823 patients, but 19 ran in China, 18 were published in Chinese, only 1 was registered, and the funnel plot was lopsided, which suggests negative trials went missing.¹¹

The best-known American trial was an NIA-funded phase II study in Alzheimer’s disease. Results by dose:

• 200 mcg twice daily: negative on the primary outcome (p=0.98).
• 400 mcg twice daily: ADAS-Cog improved by 2.27 points at week 11 (p=0.001) and 1.92 points at week 16 (p=0.07), in secondary analyses.
• Funding and safety: a HupA supplier co-funded the trial, and serious adverse events rose with dose, at 8.2%, 13.0%, and 16.2%.¹²

A 720-patient phase II/III trial against donepezil began recruiting in 2025, with primary completion due in August 2028.²²

What is the evidence in healthy adults?

Almost none. A search of PubMed, ClinicalTrials.gov, and Consensus turned up no multi-week trial of huperzine A alone in healthy adults. Three small studies exist:

• Sun 1999: 34 matched pairs of Chinese teenagers with memory complaints took 100 mcg twice daily for 4 weeks. Memory quotient reached 115 against 104 on placebo (P<0.01). It is the only positive repeated-dose result in healthy people, and no one has replicated it.⁷
• Morasch 2014: a single 100 or 200 mcg dose, 12 adults per arm. Blood acetylcholinesterase fell, but performance did not improve, apart from maintained recall at 200 mcg.⁵
• Wessinger 2021: 15 trained adults took 200 mcg before exercise. No cognitive measure differed (all p of 0.296 or higher), and exercise felt harder on HupA (6.8 against 5.7 out of 10, p=0.002).⁶

For wider context on how brain supplements are judged, see Do Supplements for Brain Health Really Work. For caregivers: huperzine A does not replace prescribed treatment, and any use belongs in a conversation with the prescriber.

Dosage, safety, and interactions

Supplements typically list 50 to 200 mcg per capsule. Alzheimer’s trials used 200 to 800 mcg per day. This page reports those figures and does not recommend a dose.

How long does it last? Half-life reports vary by study:

• About 6 hours (n=18 men).
• 11.9 hours (n=12).
• 12.1 hours (fasting bioequivalence study).

The widely quoted 10 to 14 hours traces to one 12-person study. No study has tested food timing or a cycling schedule. The case for cycling rests on animal evidence that stronger cholinesterase inhibitors reduce receptor numbers over time. In the one HupA test, rats given daily oral doses for 30 days kept the same rise in acetylcholine.

Nausea, diarrhea, sweating, dizziness, and a slower heart rate follow from the mechanism. Avoid HupA in pregnancy and breastfeeding. Four groups face interaction risk:

• Other cholinesterase inhibitors (donepezil, galantamine, rivastigmine), which can add to the effect.
• Anticholinergic drugs, which HupA can blunt.
• Beta blockers and other heart-rate-lowering drugs.
• Surgery. Cholinesterase inhibitors change neuromuscular blockade, so an anesthetist should know about use. This is class guidance, with no HupA-specific case report.

Sleep has no HupA data. Galantamine, a related drug, raised lucid dreaming from 14% on placebo to 27% at 4 mg and 42% at 8 mg in 121 people. Vivid dreams fit the drug class, but nobody has tested them for HupA.

Should huperzine A be stacked with alpha-GPC?

Stacks pair huperzine A with a choline source, on this logic: brain choline follows plasma choline, and active neurons draw choline from their own membranes.²⁴ The logic is plausible. The evidence is missing. No human trial has tested huperzine A with choline or alpha-GPC. The only trial of a cholinesterase inhibitor plus a choline donor used donepezil 10 mg with alpha-GPC 1,200 mg a day in Alzheimer’s patients with vascular injury. Interim reports favored the combination, and one investigator group ran all of them.²⁵

Quality, testing, and legal status

A 2020 analysis of 22 products listing huperzine A found only 2 within 10% of the declared amount, and 16 (73%) held unlisted compounds.¹⁶ A later review reported that 8 products labeled as plant or extract showed HupA with no other plant markers, which points to isolated or synthetic HupA. The full text of the original analysis was not available for this article, so the figures come from the published abstract.

How to choose a huperzine A product

A label that passes the decoder meets these checks:

• HupA stated in mcg, or extract mg plus a percentage.
• Source named: species, plant part, and natural or synthetic.
• Third-party testing or a certificate of analysis with HPLC results.
• No proprietary blend hiding the HupA amount.
• No hidden stimulants.
• A standardized extract or isolated HupA over whole-herb powder.

Frequently asked questions

Is huperzine A the same as Huperzia serrata?

No. Huperzine A is a single compound. Huperzia serrata is one of several club mosses that make it, and Phlegmariurus species generally hold more HupA than Huperzia species.

Is Huperzia serrata extract the same as huperzine A?

Not by default. The extract is a plant powder standardized to a percentage of HupA, often 1%. A 20 mg extract at 1% delivers 200 mcg. A label with no percentage gives no dose.

What plant does huperzine A come from?

Mainly Huperzia serrata and related club mosses, such as Phlegmariurus carinatus, which held the most HupA in a 2005 survey. HupA can also be made by chemical synthesis.

Is huperzine A natural or synthetic?

Both exist. Natural HupA is the (-) form. Synthetic material is often a 50:50 mix with about half the activity, though pure synthetic (-)-HupA also exists. Carbon-14 testing tells plant and petrochemical sources apart.

How much huperzine A is in Huperzia serrata extract?

It depends on the grade. A 1% extract holds 10 mg of HupA per gram. Raw moss holds about 0.005% to 0.025%, so a 1% extract is 40 to 200 times more concentrated.

Is huperzine A anticholinergic?

No, it does the opposite. HupA is a cholinesterase inhibitor, so it raises acetylcholine. It can add to cholinergic drugs and blunt anticholinergic ones.

Who should not take huperzine A?

People on donepezil, galantamine, rivastigmine, anticholinergic drugs, or heart-rate-lowering drugs, and anyone pregnant, breastfeeding, facing surgery, or with seizure or heart conditions. A doctor or pharmacist should review any use.

Is Huperzia serrata the same as Boswellia serrata?

No. Boswellia serrata is a resin-producing tree used for frankincense extracts. Huperzia serrata is a club moss. They share only a species epithet.

Sources

1. Ma X, Tan C, Zhu D, Gang DR. Is there a better source of huperzine A than Huperzia serrata? J Agric Food Chem 2005;53:1393-8. doi:10.1021/jf048193n (PMID 15740012)
2. McKinney M et al. Potencies and stereoselectivities of enantiomers of huperzine A. Eur J Pharmacol 1991;203:303-5. doi:10.1016/0014-2999(91)90730-e
3. Wang Y et al. [+]-Huperzine A protects against soman toxicity in guinea pigs. Neurochem Res 2011;36:2381-90. doi:10.1007/s11064-011-0564-5
4. Haigh JR et al. Protection of red blood cell acetylcholinesterase by oral huperzine A. Chem Biol Interact 2008;175:380-6. doi:10.1016/j.cbi.2008.04.033
5. Morasch KC et al. Physiological and neurobehavioral effects of cholinesterase inhibition in healthy adults. Physiol Behav 2014;138:165-72. doi:10.1016/j.physbeh.2014.09.010
6. Wessinger CM et al. Effect of huperzine A on cognitive function and perception of effort during exercise. Int J Exerc Sci 2021;14:727-41. doi:10.70252/GQBM6956
7. Sun QQ et al. Huperzine-A capsules enhance memory and learning performance in 34 pairs of matched adolescent students. Acta Pharmacol Sin 1999;20:601-3. PubMed 10678121
8. Lee G et al. Association of L-alpha glycerylphosphorylcholine with subsequent stroke risk after 10 years. JAMA Netw Open 2021;4:e2136008. doi:10.1001/jamanetworkopen.2021.36008
9. Nett RS et al. Plant carbonic anhydrase-like enzymes in neuroactive alkaloid biosynthesis. Nature 2023;624:182-91. doi:10.1038/s41586-023-06716-y
10. Li J et al. Huperzine A for Alzheimer’s disease. Cochrane Database Syst Rev, CD005592. Cochrane
11. Yang G et al. Huperzine A for Alzheimer’s disease: a systematic review and meta-analysis of randomized clinical trials. PLoS ONE 2013. PMC3781107
12. Rafii MS et al. A phase II trial of huperzine A in mild to moderate Alzheimer disease. Neurology 2011. Registry: NCT00083590
13. Felgenhauer N et al. Fir clubmoss (Lycopodium selago) tea poisoning. J Toxicol Clin Toxicol 2000;38:803-8. doi:10.1081/clt-100102396
14. DTU Fødevareinstituttet. Risk assessment of Huperzia serrata, 21 April 2016. PDF
15. Operation Supplement Safety. Huperzine A: dietary supplements for brain health. opss.org
16. Crawford et al. Analysis of 22 huperzine A products (abstract), Clin Toxicol HDS abstracts, 2020. PDF
17. Beta Analytic. Natural vs synthetic huperzine A case study. betalabservices.com
18. Goodger JQD et al. Variation in huperzine A and B in Australasian Huperzia species, 2008. JCU; Lim et al. 2010 (same group).
19. RIVM (Netherlands). Huperzine A report, 2024, reporting Ma 2005 plant-content figures.
20. NIH Office of Dietary Supplements. Dietary Supplement Label Database; sample of 308 labels parsed for this article.
21. ChEMBL compound report, huperzine A (CHEMBL395280). EMBL-EBI
22. ClinicalTrials.gov NCT07066826, huperzine A controlled-release tablets vs donepezil. Registry
23. Wong JC et al. Huperzine A provides robust and sustained protection against induced seizures. Front Pharmacol 2016. doi:10.3389/fphar.2016.00357; Supernus press release, 23 May 2024.
24. Blusztajn JK, Wurtman RJ. Choline and cholinergic neurons. Science 1983;221:614-20. doi:10.1126/science.6867732
25. Amenta F et al. ASCOMALVA trial. J Neurol Sci 2012;322:96-101. doi:10.1016/j.jns.2012.07.003
26. Kim HK et al. Choline alphoscerate in mild cognitive impairment. J Prev Alzheimers Dis 2025;12:100059. doi:10.1016/j.tjpad.2025.100059