Semax

ACTH(4-10) analog

Semax is a nootropic peptide with something almost no other research peptide has: a decades-long track record as an actual approved pharmaceutical. Developed in Russia in the 1980s and registered there since 1994, it's used as a nasal spray for cognition, stroke recovery and focus. This guide covers how Semax works, what the research shows, dosing references, its unusually clean safety profile, and status.

How Semax works

Semax is a synthetic heptapeptide derived from ACTH (4-10) — but engineered to keep the cognitive effects while removing the stress-hormone (steroidogenic) activity of its parent. Its headline mechanism is rapid upregulation of BDNF (brain-derived neurotrophic factor) and its TrkB receptor in the hippocampus — BDNF is essentially fertilizer for neurons and synapses. It also modulates dopamine and serotonin signaling and appears to act partly via melanocortin (MC4R) pathways. The result researched is sharper focus, memory and processing without classic stimulant jitter.

What the research & clinical use show

Semax is genuinely unusual: it's a registered drug in Russia, used clinically for ischemic stroke recovery, cognitive impairment, optic-nerve conditions and attention deficit. Russian clinical studies report improvements in memory, attention, processing speed and executive function — even in cognitively normal people — with effects building over days to weeks. The honest caveat for a Western audience: most of this evidence comes from Russian research and it hasn't been through the FDA process, so it's less familiar to US clinicians despite the real track record.

Semax dosing (research reference)

Semax is used intranasally. The literature references a cognitive-enhancement range of roughly 600–1,200 mcg/day using the 0.1% solution (about 50 mcg per drop); higher-strength 1% solutions are referenced for stroke/optic-nerve protocols. Because it's nasal, there's no injection or reconstitution for the standard cognitive use. This summarizes existing references for education only, not dosing advice.

Safety & side effects

Semax has one of the cleanest safety profiles in the neuropeptide space. Its decades of Russian pharmaceutical use — including in pediatric, stroke and healthy-adult populations — come with no major reported adverse effects and no known addictive potential or withdrawal. That's a strong record, though as always this is educational information, not medical advice, and product quality on the research market varies.

Related compounds

Semax is often discussed alongside its sister peptide Selank (an anxiolytic-leaning neuropeptide from the same Russian research lineage) and the broader nootropic family. It's a go-to for focus and neuroprotection rather than raw stimulation.

Legal & regulatory status

Semax is an approved pharmaceutical in Russia but is not FDA-approved in the US, where it's sold as a research compound (research use only). Follow the laws that apply to you.

Where to get Semax

Semax is sold in 2 forms. They are not interchangeable — the dose and the route differ, so pick the one this page describes unless you know why you want another.

Semax reconstitution calculator

The evidence for Semax

Graded by what exists behind each claim.

Human clinical evidence

• The human record is Soviet and post-Soviet clinical work — real patients and real endpoints, published in Russian and rarely replicated to Western standards.

📊 Correlative data

• Registered as a medicine in Russia for stroke, transient ischemic attack and optic nerve conditions, with Russian clinical trials behind those registrations. It is on the Russian Vital and Essential Drugs list, which is a real regulatory position even though no Western agency has assessed it.
• Widely used intranasally for focus and recovery. The reported experience is consistent — clean, non-stimulant alertness without a comedown — which is unusual enough in the anecdotal record to be worth noting.

🧪 Theoretical / extrapolated

• A fragment of ACTH (4–10) with the corticotropic activity removed, so it does not raise cortisol. It increases BDNF and NGF expression in rodent brain, which is the proposed mechanism for both the acute focus effect and the neurorestorative claims.
• The BDNF mechanism predicts a slow-building effect over weeks alongside the acute one, and it is why the Russian stroke protocols run for courses rather than single doses.

How to read the Soviet clinical series →

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What Semax actually does

Semax is Met-Glu-His-Phe-Pro-Gly-Pro, printed in those seven residues in the abstract of the paper that measured its effect on human serum enzymes Kost 2001. Its first four residues are ACTH(4–7); the Pro-Gly-Pro is synthetic and appears nowhere in proopiomelanocortin. The label ‘ACTH(4–10) analog’ describes its lineage, not its structure — both names are used in the literature, one paper titling it ACTH(4-10) analog Glazova 2021 and another ACTH(4-7)PGP Filippenkov 2020.

The substitution is the whole molecule, and here is what it actually deletes. Native ACTH(4–10) is Met-Glu-His-Phe-Arg-Trp-Gly. Semax replaces Arg-Trp-Gly with Pro-Gly-Pro. The melanocortin receptor recognition motif — the core pharmacophore used as a comparator peptide in receptor experiments — is His-Phe-Arg-Trp Getting 2003. Semax keeps His-Phe and deletes Arg-Trp. So ‘the corticotropic activity has been removed’ is not a marketing hedge: it is the deletion of half of a four-residue receptor-binding motif, and it is checkable by anyone who can read two sequences side by side. The other POMC fragment in this catalog, KPV, keeps none of those four.

The copper site nobody mentions. A peptide with a free N-terminal amine and a histidine in the third position carries the canonical Xaa-Xaa-His motif, the square-planar Cu(II) chelation architecture found in serum albumin's Asp-Ala-His N-terminus and in the copper-carrying cosmetic tripeptide GHK. Semax is Met-Glu-His… — it has it. And the chemistry behaves exactly as the motif predicts: Semax forms a stable complex with Cu(2+) and prevents formation of the Abeta:Cu(2+) complex Sciacca 2022, and it will strip Cu(II) out of Cu(II)-Abeta, silence its redox cycling, cut the associated reactive oxygen species and protect SH-SY5Y cells from copper-catalyzed oxidative damage Tomasello 2025. This is the most mechanistically specific thing published about the molecule and it appears on no page that sells it.

The methionine is the vial's weak point. Methionine is the most oxidation-prone residue in peptide chemistry, and it sits at position 1 with a free N-terminus. Air, light and time convert it to methionine sulfoxide, which adds 16 Da and changes the N-terminal region that carries both the copper site and whatever receptor interaction exists. Nothing on a vendor page tells a buyer which of the two molecules is in the tube.

Charge and mass, and the contrast with its usual partner. 813.9 Da, formula C37H51N9O10S, computed isoelectric point 5.36 and net charge about −0.9 at blood pH. Selank, the peptide it is stacked with, is about +1.9. They are opposite in sign, which is why they do not share handling, mucosal binding or purification behavior.

No named receptor. There is no cloned receptor for Semax. The nearest mechanistic anchor is binding work in isolated rat brain plasma membranes, where Semax and its sibling ACTH-fragment-PGP peptides each altered [3H]GABA binding in their own way and differently at the high- and low-affinity sites, while acute restraint stress lowered binding at the low-affinity site and peptide pre-administration did not prevent that Vyunova 2023. The negative result belongs on the page as much as the positive ones.

Cell, rodent, human — and where it stops

In cells. Rat glial cell cultures respond to Semax with rapid induction of neurotrophin mRNAs Shadrina 2001, and the copper work above is cell-level too: SH-SY5Y neuroblastoma cells were protected against copper-catalyzed Abeta oxidation Tomasello 2025.

In rodents, with the route and the dose. The kinetics study is the one worth memorizing: tritium-labeled Semax at 56 Ci/mmol, given intranasally at 50 µg/kg in 20 µL to white rats of 200–250 g. Two minutes later, 0.093% of the administered radioactivity per gram was in brain, and 80% of that was still intact Semax Shevchenko 2006. In a transient middle cerebral artery occlusion model, Semax changed 394 genes by more than 1.5-fold at 24 hours against saline, suppressing inflammatory genes and activating neurotransmission genes — the mirror image of what the ischemia itself did Filippenkov 2020. In a developmental model, rat pups given fluvoxamine on postnatal days 1–14 and then Semax on days 15–28 had less anxiety-like behavior, better learning and normalized brain monoamines in the second month of life Glazova 2021.

In humans, and this is a real clinical dataset. 110 patients after ischemic stroke — 43 men, 67 women, mean age 58.0 ± 9.7 — split into early (89 ± 9 days) and late (214 ± 22 days) rehabilitation groups, with and without semax. The regimen was two courses of 6000 µg/day for 10 days, 20 days apart. Plasma BDNF rose and stayed high, Barthel index recovery accelerated, and BDNF correlated with Barthel score Gusev 2018. Separately, 52 healthy volunteers had resting-state fMRI before injection and at 5 and 20 minutes after Semax, Selank or placebo, with changes in right amygdala to right temporal cortex connectivity Panikratova 2020.

The obstacles, and the first one is arithmetic.

• The dose gap is 7.5- to 30-fold. The only human trial with a stated regimen used 6000 µg a day Gusev 2018. This site's protocol is 200–800 µg. Nothing about a stroke-rehabilitation result at 6000 µg/day transfers automatically to a tenth of that dose, and no dose-ranging study exists between the two.
• The population gap. The human efficacy data are in people who have already had a cerebrovascular event, with Barthel index and plasma BDNF as endpoints. Neither is a measure of focus in a healthy adult. The one healthy-volunteer study measured resting connectivity, not performance Panikratova 2020.
• The delivery gap. The rat kinetics used a pipetted 20 µL intranasal volume Shevchenko 2006. A self-administered spray deposits rather than delivers: the fraction swallowed, the fraction on the anterior septum and the fraction reaching olfactory epithelium vary by head position and technique, and none of that variance exists in the rat number. Where a nasal product is labeled as a percentage solution rather than micrograms per actuation, the delivered dose has to be computed from the metered volume, not read off the box.
• The comparator for a healthy user is not another peptide. For attention and working memory in healthy adults, sleep extension, aerobic exercise and caffeine all have larger and far better replicated effect sizes than anything published here.
• Every human study is single-country and Russian-language, and none appears in a Western trial registry.

Semax pharmacokinetics — how much of it actually gets in

What degrades it, and into what. Semax undergoes rapid enzymatic degradation, and in biological samples after intranasal dosing the tripeptide Pro-Gly-Pro predominates over intact Semax and its other metabolites Shevchenko 2006. So the Pro-Gly-Pro tail is doing two jobs, and only the first is ever mentioned: it slows proteolysis of the intact peptide, and it is released as a compound that is independently active. In primary neuroglial culture, PGP at 100 µM given before a scratch injury prevented the calcium rise and the mitochondrial depolarization that follow it, and at 30 µM for three to six days it reduced glutamate-induced calcium disturbance and cell death, raised intracellular BDNF and lowered extracellular NSE Bakaeva 2024. A person dosing Semax is, within minutes, also dosing PGP.

The number that anchors everything. 0.093% of the administered radioactivity per gram of brain at two minutes, from a 50 µg/kg intranasal dose, 80% of it intact Shevchenko 2006. Scale that: a 70 kg adult taking 800 µg is at about 11 µg/kg, roughly a fifth of the rat's dose per kilogram before any correction for the far larger surface of a rat's olfactory epithelium relative to its body. This is why the route matters more here than the milligram count.

The oral barrier. Nobody sells this to be swallowed, and the reasons are specific rather than generic: gastric acid, pancreatic proteases and brush-border peptidases, a 7-residue length that is outside the di/tripeptide range of the PepT1 carrier, and a methionine that oxidizes. No oral bioavailability figure has ever been published for it in any species.

The injectable comparator. Subcutaneous injection removes the nasal deposition variance and the gut entirely, at the cost of handing the whole dose to plasma peptidases at once. What does not exist is the experiment that would let anyone choose between the routes: no published study has compared subcutaneous and intranasal Semax for brain exposure in the same animals, so the widely repeated claim that the injection ‘hits harder’ has no measurement behind it in either direction.

What would have to be true, and how you would know it was not

Three predictions, one of which is a copper experiment nobody has proposed before.

1. Morning cortisol will not rise, and that is also an identity test. Semax lacks Arg-Trp, half the His-Phe-Arg-Trp melanocortin pharmacophore Getting 2003, and no melanocortin-receptor agonism has been reported for it. Draw AM cortisol at baseline and at 4 weeks and expect no change. A rise would mean either the mechanism claim is wrong or the vial contains a longer, genuinely corticotropic ACTH fragment — which is the cheapest way a buyer can test the identity of what they bought.

2. Serum copper and ceruloplasmin will not move, and here is the arithmetic. Semax is a Xaa-Xaa-His copper chelator that strips Cu(II) from Abeta in vitro Sciacca 2022 Tomasello 2025. At 800 µg a day of an 813.9 Da molecule that is about 0.98 µmol/day, against roughly 33–66 µmol of copper circulating in plasma at any moment. One to three per cent of the plasma pool, from a molecule that is degraded within minutes, should not shift serum copper or ceruloplasmin at 8 weeks. If either did move, the chelation chemistry is far more systemic than anyone has assumed and that would be a genuinely new finding rather than a side effect.

3. Plasma BDNF will not move at this dose, and that is the prediction that cuts against the product. The human BDNF result exists, but it is in stroke patients at 6000 µg/day Gusev 2018. At 200–800 µg/day in a healthy person the expectation is no change outside assay noise — and before anyone runs it, note that circulating BDNF is dominated by release from platelets, so serum and plasma give different numbers and a change of tube type will move the result more than this compound will. Anyone testing it must use the same matrix, the same lab and the same time of day at both draws.

What nobody has tested yet

Six experiments, and the first one costs less than a vial.

1. Nobody has checked whether the vial contains Semax or Semax methionine-sulfoxide. A single LC-MS run on a month-old opened nasal bottle would tell every user of this compound something nobody currently knows. The oxidized form is a different molecule at the exact end of the peptide that carries the copper site.

2. Nobody has given Pro-Gly-Pro to a person. It is the dominant species in tissue after an intranasal Semax dose Shevchenko 2006 and it is independently neuroprotective in culture Bakaeva 2024. If the tripeptide carries part of the effect, then the correct product is three residues long and costs a fraction of what the heptapeptide costs. The comparison has never been run in any species side by side at equimolar dose.

3. The copper question in a living animal. Does chronic intranasal Semax change brain or serum copper handling? The chemistry is published twice over Sciacca 2022 Tomasello 2025; the animal experiment is a four-week dosing study with tissue copper by ICP-MS, and it has not been done.

4. The His-to-Ala experiment. Substituting the histidine at position 3 abolishes the Xaa-Xaa-His copper site while leaving the rest of the molecule intact. Run the analog in the ischemia model that Semax already works in Filippenkov 2020 and the copper mechanism either accounts for part of the neuroprotection or it does not. This is a first-year experiment on a compound that has been sold for thirty years.

5. Nasal against subcutaneous, brain exposure, same animals. The community argues about this constantly and no one has measured it.

6. A dose-response in healthy people. The human evidence sits at 6000 µg/day in stroke Gusev 2018 and at a single injection under an MRI scanner in health Panikratova 2020. Nothing exists in between. A three-arm study at 300, 1000 and 3000 µg/day with a working-memory battery would be the first cognitive dose-response this compound has ever had.

Semax — its own safety story, not its class's

The class safety block on this page is written for injectable nootropics as a group. Four things here belong to this molecule alone.

1. The most popular route is the one the class block never mentions. Most people take this intranasally. That puts a repeated, self-administered chemical onto ciliated respiratory epithelium several times a day for weeks, and there is no published study of nasal mucosal tolerance for this peptide at any dose or duration. Absence of reported irritation is not the same as an examined nose.

2. Product identity is a safety issue, not a quality issue. The methionine oxidizes; nothing on a label distinguishes the peptide from its sulfoxide; and where a product is specified as a percentage solution the delivered microgram dose per actuation is a calculation the buyer has to do.

3. This is a registered stroke medicine being used off-label at a tenth of the studied dose. In the country where it is approved it is given at 6000 µg/day, in ten-day courses, to people under medical supervision after a cerebrovascular event Gusev 2018. Taking one tenth of that for focus is a different act with a different risk calculation, and the most consequential version of the risk is substitution: new neurological symptoms — sudden weakness, speech change, facial droop, worst-ever headache — are an emergency, and the presence of a neuroprotective peptide in the fridge changes nothing about that.

4. The two datasets that look like safety data and are not. In mouse embryonic stem cells, Semax at 10 and 0.1 µM increased survival under serum deprivation and had little effect on formation of neuronal precursors; the authors read their results as showing no toxic effect during the embryonic and fetal period Kobylyanskii 2017. That is a dish, not a pregnancy, and it is the only developmental datum that exists. Separately, Semax reversed the behavioral and monoamine consequences of neonatal SSRI exposure in rats Glazova 2021 — which shows it acts on the same monoamine systems an antidepressant does. No interaction study exists in an adult on an SSRI, and that is precisely the combination most likely to be running in real life.

Sources read for this page

• Kost NV, Sokolov OY, et al. Semax and selank inhibit the enkephalin-degrading enzymes from human serum. Bioorganicheskaia Khimiia 2001 · PMID 11443939
• Shevchenko KV, et al. Kinetics of Semax penetration into the brain and blood of rats after its intranasal administration. Bioorganicheskaia Khimiia 2006 · PMID 16523722
• Gusev EI, et al. The efficacy of semax in the tretament of patients at different stages of ischemic stroke. Zhurnal Nevrologii i Psikhiatrii im S S Korsakova 2018 · PMID 29798983
• Filippenkov IB, et al. Novel Insights into the Protective Properties of ACTH(4-7)PGP (Semax) Peptide at the Transcriptome Level Following Cerebral Ischaemia-Reperfusion in Rats. Genes (Basel) 2020 · PMID 32580520
• Glazova NY, et al. Semax, synthetic ACTH(4-10) analogue, attenuates behavioural and neurochemical alterations following early-life fluvoxamine exposure in white rats. Neuropeptides 2021 · PMID 33418449
• Vyunova TV, et al. Synthetic corticotropins and the GABA-receptor system: Direct and delayed effects. Chemical Biology and Drug Design 2023 · PMID 36828803
• Bakaeva Z, et al. Regulatory Peptide Pro-Gly-Pro Accelerates Neuroregeneration of Primary Neuroglial Culture after Mechanical Injury in Scratch Test. International Journal of Molecular Sciences 2024 · PMID 39456669
• Sciacca MFM, et al. Semax, a Synthetic Regulatory Peptide, Affects Copper-Induced Aβ Aggregation and Amyloid Formation in Artificial Membrane Models. ACS Chemical Neuroscience 2022 · PMID 35080861
• Tomasello MF, et al. Semax, a Copper Chelator Peptide, Decreases the Cu(II)-Catalyzed ROS Production and Cytotoxicity of Aβ by Metal Ion Stripping and Redox Silencing. Bioinorganic Chemistry and Applications 2025 · PMID 40496623
• Shadrina MI, Dolotov OV, Grivennikov IA, Slominsky PA, Andreeva LA, Inozemtseva LS, Limborska SA, Myasoedov NF. Rapid induction of neurotrophin mRNAs in rat glial cell cultures by Semax, an adrenocorticotropic hormone analog. Neuroscience Letters 2001 · PMID 11457573
• Panikratova YR, et al. Functional Connectomic Approach to Studying Selank and Semax Effects. Doklady Biological Sciences 2020 · PMID 32342318
• Kobylyanskii AG, et al. Studying the Toxic Effects of Some Biologically Active Peptides on the Model of Mouse Embryonic Stem Cells. Bulletin of Experimental Biology and Medicine 2017 · PMID 29063333
• Getting SJ, et al. Dissection of the anti-inflammatory effect of the core and C-terminal (KPV) alpha-melanocyte-stimulating hormone peptides. Journal of Pharmacology and Experimental Therapeutics 2003 · PMID 12750433

Semax — safety, predicted from mechanism

Semax — interference & stacking

Bloodwork to run alongside Semax

Baseline first, then again at 8–12 weeks.

The Brain Fog & Cognition panel covers these in one order — 12 markers, $233.06 with the discount applied.

Check results you already have → · All 103 markers A–Z

Semax — frequently asked questions

References & further reading

1. Semax: mechanism, effects & research studies (Peptpedia)

Semax inside a finished plan

One arm of 1 Protocol Blueprint, free to read in full.

What Semax is used for

Semax appears under 1 goal in the goal router.

Related Cognitive & Mood compounds

← Explore the full Protocol Vault

↑ Back to on this page