Semax (ACTH(4-7)-Pro-Gly-Pro)
Semax: What The Studies Actually Say
It is sold as a brain booster. The brain-growth results come from rats. The human studies are small, mostly in Russian and nearly all from the research network that made it, and none settles whether it sharpens a healthy mind. Here is what the research actually found.
Updated 22 September 2026
Before you read this
This page explains what published studies found. That is all it is. It is not medical advice, and it is not instructions for using anything. Nothing here is meant to find, treat, cure or stop any illness. The compounds discussed on this site, retatrutide included, are not approved by the FDA for use in people. Research compounds are sold for laboratory use only — not for use in humans or animals. A study tells you what happened to the people in it. It does not tell you what would happen to you. What you do with what you read here is your responsibility, so talk to a doctor first. This page is published by Ascend Vials, which sells research compounds.
Read the full guide
Confirm you have read the notice above and enter an email address. We will send you a link back to this guide, and sometimes an email when we publish a new one. You can unsubscribe any time.
You have probably heard that Semax raises BDNF, the brain's growth protein, and sharpens your focus. The BDNF result comes from rats. The focus claim has no solid human study behind it.
Here is what the research actually found. Every claim has a link. Read them yourself.
The 30-second version
- Semax is seven amino acids: a piece of the stress hormone ACTH with three more added on. It was designed in Moscow.
- It is a registered medicine in Russia, sold as nasal drops. It is not approved in the US, or anywhere else we could verify. A July 2026 FDA advisory vote is not an approval, and it is not final.
- The human evidence is small and early. Many studies are in stroke patients. Most are in Russian, and nearly all come from the research network that made it. Most do not describe random assignment or blinding. None has been repeated by an independent group.
- The famous brain-growth result, BDNF, comes from rats. In people it has only been measured in the blood.
- In healthy people, the two brain-scan studies measured brain activity, not memory or focus.
- FDA says there is not enough information to describe its safety. In rodents it amplified amphetamine's effect on dopamine and made blood clots smaller.
- Nothing sold online has been shown to match what the studies used. No study has checked.
What it is
Semax is a lab-made peptide seven amino acids long: Met-Glu-His-Phe-Pro-Gly-Pro. The first four are a piece of ACTH, a stress hormone. The last three, Pro-Gly-Pro, were added on. Citing the developers, FDA says it has none of ACTH's hormone effects on the adrenal glands or skin color. FDA briefing, 2026
It was designed in Moscow, and its developers reviewed their first 15 years of work on it in 1997. Source In Russia it is sold as nasal drops. In the US, wellness clinics and online stores sell it as nasal sprays and injections. FDA briefing, 2026
The human studies are small and scattered
| People | Group that got nothing extra? | What it found | |
|---|---|---|---|
| Stroke recovery, 2018 | 110 | No Semax; no placebo described | Blood BDNF up, faster recovery. No numbers given |
| Early stroke, 1997 | 30 vs 80 | Usual care only | "Some influence" on recovery. No numbers given |
| Stroke, pooled analysis, 2018 | 3 studies, 181 | Placebo | Better stroke scores at 21 days |
| Poor blood flow to the brain, 2005 | 187 | Not described | Authors report improvement. No numbers given |
| Motor neuron disease, 2007 | 27 | No | No change in muscle or disease scores |
| Stomach ulcers, 2002 | 32 | Usual ulcer drugs only | Healed by day 14: 89.5% vs 30.8% |
| Pain, 1996 | 37 | No | Authors: no pain relief "by itself" |
| Psoriasis and cholesterol, 2015 | 118 | Usual care only | Blood fats improved. No numbers given |
| Brain injury from lack of oxygen, 1999 | 73 | No | Authors report memory help. No numbers given |
| Optic nerve disease, 2000 | Not stated | Usual care only | Authors report better vision. No numbers given |
| Healthy adults, brain scans, 2018 and 2020 | 24 and 52 | Placebo | Brain networks changed. No thinking tests |
Read the whole table. "No numbers given" means the published summary reports a result without the figures behind it. Many of these papers are in Russian. There are too few comparable numbers for an honest chart, so there is none.
Sources: 2018 stroke · 1997 stroke · pooled analysis · blood flow · motor neuron · ulcers · pain · psoriasis · oxygen · optic nerve · scans 2018 · 2020
Stroke. In 110 stroke survivors, those given Semax had blood BDNF rise and stay high for the 5-month study. Semax and high BDNF went with faster recovery of everyday function. The summary gives no effect sizes and describes no placebo or random assignment. Zh Nevrol Psikhiatr, 2018
The one pooled analysis found 364 articles, but only 3 studies, with 181 people, met its rules. Against placebo, stroke-severity scores were better at 21 days in mild, moderate and severe strokes. Look closer and it gets thinner. Each severity group pooled at most two studies. In places the results disagreed hugely between studies. Nothing is reported at 3 months, the usual checkpoint in stroke trials. The authors themselves call for a large double-blind trial. Bulletin of Rehabilitation Medicine, 2018
The result that went the wrong way. In 27 people with motor neuron disease, the family that includes ALS, Semax did not change muscle nerve damage or disease scores. Mood and motivation improved. There was no comparison group. Zh Nevrol Psikhiatr, 2007
Pain. In 37 adults given one dose, with no comparison group, migraine stopped in 4 of 12. Trigeminal neuralgia, a facial nerve pain, did not change at all. The authors concluded that Semax "does not exhibit analgesic activity by itself", and FDA judged the study not enough to show it works. FDA briefing, 2026 Keep that in mind when you read about the July 2026 vote below.
Ulcers. This result looks the strongest, but it is a two-page report on 32 people with no random assignment or blinding described. The authors ask for more studies. Bull Exp Biol Med, 2002
The brain-boost claim has not been properly tested in people
Russia's essential-medicines list once filed Semax under nootropics, drugs meant to sharpen thinking. ConsultantPlus Here is what exists in healthy people:
- 24 adults, one dose. Brain scans showed a front part of a resting brain network was larger with Semax than with placebo. No memory or thinking test was done. Bull Exp Biol Med, 2018
- 52 people, one dose. Brain connections differed between the groups. No behavior was measured. FDA notes the Semax volunteers may be the same people as in the 2018 study. Dokl Biol Sci, 2020 · FDA briefing, 2026
- 19 healthy men, 1996. FDA mentions it only in passing and did not weigh it as evidence that Semax works. FDA briefing, 2026
A brain scan that looks different is not a sharper mind. Whether Semax improves memory, attention or focus in a healthy person is still an open question.
The BDNF story comes from rats
BDNF is a protein that helps brain cells grow and connect. It is Semax's big selling point, and the brain results come from rodents. FDA says Semax works "through poorly understood mechanisms of action". FDA briefing, 2026
- BDNF. In rats, one nasal dose raised BDNF in the hippocampus, the brain's memory center, up to 1.4-fold. Treated rats did better on a learning test. Brain Res, 2006 Not every change went one way: in another study, a growth-factor gene switched down in the frontal cortex. Neurosci Lett, 2007
- Getting in. Two minutes after a nasal dose, each gram of rat brain held 0.093% of it, mostly intact. It then breaks down fast, largely into Pro-Gly-Pro, a piece with effects of its own. Bioorg Khim, 2006 · Cell Mol Neurobiol, 2010
- Stroke. After an experimental stroke, a week of Semax left the damaged area about 20% smaller than in rats given a blank treatment, and treated rats kept a learned memory. FDA briefing, 2026 · Bull Exp Biol Med, 2006
- Other models. It reduced signs of stress in rats, and helped mice bred to model Alzheimer's and mice with spinal cord injuries. Stress · Alzheimer's model · Spinal cord
So the honest line is this: Semax does many things in rodents. Nobody has shown it does them inside a human brain. No study has even measured how the human body absorbs or clears it. FDA briefing, 2026
Safety: mostly a blank page
FDA's 2026 review found "insufficient clinical information to characterize the safety profile" of Semax. FDA found no safety data for injection under the skin, and no animal studies of poisoning, repeated doses, DNA damage, reproduction or long-term cancer risk. FDA briefing, 2026
Amphetamine. On its own, Semax did not change dopamine, a brain reward chemical. Neurochem Res, 2005 But in mice, amphetamine alone raised movement to 182% of normal, and amphetamine plus Semax pushed it to 261%. Eksp Klin Farmakol, 2004 FDA called the dopamine finding "concerning", because that kind of rise is typical of drugs of abuse such as cocaine. No study has tested whether Semax itself is addictive. FDA briefing, 2026
Bleeding. In rats, 5 days of nasal Semax made the blood better at dissolving clots, and clots came out smaller. Bull Exp Biol Med, 2001 FDA says this raises "concern about the risk of bleeding", especially with other drugs that raise bleeding risk. FDA briefing, 2026
Brain electrical activity. In 73 patients with brain injury from lack of oxygen, some showed bursts of abnormal activity on a brain-wave test after a Semax injection. Anesteziol Reanimatol, 1999
Immune reactions. FDA warns that compounded Semax "may pose risk for immunogenicity", because peptides can clump and carry impurities. FDA found no clinical study of this. FDA · FDA briefing, 2026
"Well tolerated." The developers' papers say this. One reports a "minor percent of side-effects" with no rates. Zh Nevrol Psikhiatr, 2005 FDA found that most human studies did not discuss safety at all. None reports side-effect rates collected systematically. FDA briefing, 2026 Well tolerated is a claim. A side-effect rate is a number.
Who did the research
Look at the author lists in the sources below. Nearly all of the research comes from one network of Russian research groups, the one that developed Semax, and much of it is published in Russian. No human result has been repeated by an independent group, and ClinicalTrials.gov, the main public trial registry, lists zero Semax trials. ClinicalTrials.gov
None of that makes the results wrong. It means almost nobody outside the group has checked them.
Is it approved?
Semax is a registered medicine in Russia. We could not verify an approval anywhere else.
- Russia: registered and sold as nasal drops. FDA briefing, 2026 It was on Russia's essential medicines list, under "other psychostimulants and nootropics". That 2019 order is now marked no longer in force, and we could not confirm that the new list includes Semax. ConsultantPlus
- United States: not FDA-approved, and not an ingredient of any approved drug. FDA's own staff reviewed it in 2026 and recommended against letting pharmacies compound it. FDA briefing, 2026 FDA once flagged it as a possible significant safety risk, and that warning is still on its page. FDA
- The July 2026 vote: an FDA advisory committee voted 8 to 5 to recommend adding it anyway, for migraine, poor blood flow to the brain and trigeminal neuralgia. The vote does not bind FDA, and STAT reports any decision would first come as a proposed rule open for public comment. STAT News · NCPA We found no such rule naming Semax as of 22 September 2026. Federal Register search
- Elsewhere: FDA found no Semax standard in the European, Japanese or International pharmacopoeias, the official drug-quality rulebooks. FDA briefing, 2026
What is still unknown
- How the human body absorbs and clears it, and how much reaches the brain.
- Whether it works in a large, randomized, double-blind trial against a placebo.
- Whether stroke patients do better at 3 months, the standard checkpoint.
- Whether it improves memory or focus in healthy people, or raises BDNF in the human brain.
- Long-term safety, cancer risk, and whether injecting it under the skin is safe.
- Whether it raises bleeding risk in people, or carries a risk of abuse.
- What it actually binds to in the body.
What is in the bottle is not what was studied
One more thing.
The human studies on this page mostly gave Semax as nasal drops. FDA notes that the Semax in one clinical study was made at a Russian Academy of Sciences lab. FDA briefing, 2026
No study has ever shown that a product sold online as "Semax" is the same thing. Not the same purity. Not the same strength. Not even proof it is the same molecule. FDA found several salts and derivatives, including different active molecules, sold under the one name. None of the US websites it reviewed said which form they sold. FDA briefing, 2026 Belgian public-health scientists found Semax in suspicious preparations. They call it and a similar peptide "potentially dangerous molecules" that testing labs must be able to identify. Source
FDA's side-effect database held one Semax case through December 2025. A woman had eye pain and burning after using Semax nasal drops bought online. She reported being hospitalized, and the pain had not gone away a year later. FDA briefing, 2026 One report cannot prove the drops caused it.
So take every number on this page for what it is. It describes what happened in those studies. It describes nothing else.
How to check a number yourself
You are going to see more claims about Semax. In a reel. In a comment. On somebody's store page.
Here is how to tell a real one from a marketing one. Five questions. About ten seconds each.
1. Is there a link to a study?
No link means it is a claim, not a result. Search "semax" on pubmed.gov yourself, then check whether the summary gives numbers. Many Semax summaries do not.
2. Was it people or rats?
The BDNF numbers people quote come from rats. In people, BDNF was only measured in blood.
3. Was there a group that got nothing extra?
Most Semax studies compared it with usual care, or with nothing. Without a placebo group, nobody can tell how much of the recovery was the drug and how much would have happened anyway.
4. Did they measure the thing being promised?
The healthy-people studies measured brain scans, not memory. The pooled stroke analysis stopped at 21 days.
5. Who ran it, and has anyone repeated it?
Here, nearly always the network that developed it. No independent group has repeated a human result.
Now run those five on this page
Every number above links to its paper. You were told which results come from rats, which summaries give no numbers, and which studies had no comparison group. The focus claim has no numbers from a proper human trial, because no such trial has been published.
That is the whole difference: a number you can check, and a number you are asked to believe.
This page has a date on it, near the top. If a proper, independent trial of Semax is ever published, this page changes.
Read the studies yourself
- FDA briefing, 2026 — safety, status, the pain and healthy-men studies
- Developers' review, 1997
- Stroke: 2018 · 1997 · pooled analysis, 2018 · blood flow, 2005
- Other patients: motor neuron disease · ulcers · psoriasis · lack of oxygen · optic nerve
- Healthy adults: 2018 · 2020
- Rats: BDNF · genes · reaching the brain · stroke · Pro-Gly-Pro · stress
- Mice: Alzheimer's model · spinal cord
- Safety signals: dopamine, 2005 · amphetamine, 2004 · clotting, 2001
- Semax in suspicious preparations, Belgium
- FDA compounding safety-risk list
- The vote: STAT News · NCPA · Federal Register search
- Russia's essential-medicines order, 2019
- ClinicalTrials.gov search
Anything above about amounts describes what a study gave to animals or enrolled patients. This is not advice, a plan, or instructions. Semax is not an approved medicine in the United States. Research compounds are not for human or animal use.