Selank was tested in people for anxiety and Semax for stroke recovery, both in small Russian studies; neither has been tested for focus in healthy adults. Neither is FDA-approved as of October 2026.
If the question is anxiety, Selank is the one with patient data: two Russian studies of 62 and 70 people. If the question is the brain after a stroke, Semax is the one with patient data: Russian studies of 110 and 187 people. For focus or memory in a healthy person, neither has a trial, and Semax's own Russian label lists anxiety disorders as a reason not to use it.
Every claim on this page is labeled by what was tested: in people, in animals or cells, or a regulatory fact. How we grade evidence: editorial policy.
| Item | Selank | Semax |
|---|---|---|
| What it is | Seven-amino-acid copy of tuftsin, an immune peptide, with a Pro-Gly-Pro tail | Seven-amino-acid copy of a piece of ACTH, a stress hormone, with the same Pro-Gly-Pro tail |
| Tested in people for | Anxiety and neurasthenia | Stroke and poor blood flow to the brain |
| Evidence label | Tested in people (small, not placebo-controlled) | Tested in people (small, not placebo-controlled) |
| FDA status | Not approved; cannot be compounded | Not approved; advisory vote 8-5 in favor, no FDA decision |
| WADA 2026 | Not named on the list | Not named on the list |
| Last checked | October 15, 2026 | October 15, 2026 |
New trials, FDA changes and side-effect reports, written in plain English. Every result says whether it was tested in people or in animals.
Each row names its evidence type. Every study named here is linked in the sources list at the bottom of the page.
| Selank | Semax | Evidence type | |
|---|---|---|---|
| Registered in Russia for | Anxiety states, neurasthenia, adjustment and stress-related disorders in adults (0.15% nose drops, sold without a prescription) | 1% drops: the first days after an ischemic stroke. 0.1% drops: memory and attention problems after stroke or head injury, optic nerve disease, and children 5 and older with minimal brain dysfunction (prescription only) | Regulatory fact |
| Largest patient study | 70 people with anxiety disorders: phenazepam alone vs phenazepam plus Selank (Medvedev, 2015) | 187 people with chronic poor blood flow to the brain (Gusev, 2005) | Tested in people |
| Headline human result | Anxiety scores fell about as much as with medazepam, a benzodiazepine, in 62 patients (Zozulia, 2008) | Higher blood BDNF and better scores on daily-living tasks in 110 stroke patients (Gusev, 2018) | Tested in people |
| What it was compared against | Another anxiety drug (medazepam or phenazepam), not placebo | Standard stroke care, not placebo | Tested in people |
| Tested in healthy people | One brain-scan study of 52 people; scans only, no anxiety score | Two brain-scan studies (24 and 52 people); scans only, no memory or focus test | Tested in people |
| What it does in the lab | Changed 45 of 84 nerve-signaling genes in rat brain within an hour, including GABA genes | Raised BDNF, a nerve growth protein, in rat forebrain within 3 hours | Tested in animals or cells |
| Shared effect | Slows the blood enzymes that break down enkephalins, the body's own opioid peptides | Same effect, about twice as strong in the same test (Kost, 2001) | Tested in animals or cells |
| Side effects on the Russian label | Allergic reactions and a bitter taste, frequency unknown | Mild irritation inside the nose with long use | Regulatory fact |
| Who should not use it (Russian label) | Pregnancy, breastfeeding, under 18 | Pregnancy, breastfeeding, anxiety disorders, acute psychosis, a history of seizures | Regulatory fact |
| Human blood-level data | Manufacturer label only: 92.8% absorbed through the nose | None: FDA found no human pharmacokinetic study by any route (2026) | Regulatory fact |
| US status, October 2026 | Not FDA-approved. FDA lists its compounding nomination as withdrawn, so pharmacies cannot use it | Not FDA-approved. FDA staff argued against compounding it; an advisory committee voted 8-5 in favor; no FDA rule yet | Regulatory fact |
They are built from different starting molecules and were tested for different conditions. Selank is a copy of tuftsin, a piece of an antibody protein, and was tested in people with anxiety. Semax is a copy of a fragment of ACTH, a stress hormone, and was tested in people after a stroke.
Both are seven amino acids long and end in the same three, Pro-Gly-Pro, which slows their breakdown in blood. Both came out of the same Moscow group: Nikolai Myasoedov of the Institute of Molecular Genetics is an author on most published studies of each.
They share one measured effect. In a 2001 lab test on human serum, both slowed the enzymes that break down enkephalins, the body's own opioid peptides, and Semax did this at about half the concentration Selank needed (tested in cells). Beyond that, the lab work points in different directions: Selank changed GABA-related genes in rat brain, and Semax raised BDNF, a nerve growth protein, in rat brain (tested in animals).
Selank is the one tested for anxiety. Semax has no anxiety trial, and its Russian label lists anxiety disorders as a reason not to use it.
Tested in people: in a 2008 study of 62 patients with generalized anxiety disorder or neurasthenia, 30 took Selank and 32 took medazepam, a benzodiazepine. Anxiety scores fell by about the same amount in both groups, and the Selank group also reported less fatigue (Zozulia, 2008).
Tested in people: in a 2015 study of 70 patients with anxiety disorders, 40 took phenazepam plus Selank and 30 took phenazepam alone. The phenazepam worked sooner with Selank added, and the group had less sedation, fewer memory and attention problems and fewer sexual side effects (Medvedev, 2015).
Neither study had a placebo group, both were run in Russia, and neither has been repeated elsewhere. They show Selank did about as well as an existing anxiety drug in those patients. They do not show how much of the effect a placebo would have produced.
Neither has been tested for focus or memory in healthy adults. Semax is registered in Russia for memory and attention problems after stroke or head injury, which is a different group of people.
The human Semax data come from patients. In 110 stroke survivors, the groups given Semax had higher BDNF in their blood and recovered more on a daily-living scale, with a smaller effect on muscle weakness (Gusev, 2018). In 1997, 30 acute stroke patients given Semax on top of standard care recovered motor function faster than 80 patients on standard care alone. Those groups were not described as randomized or blinded (tested in people).
The only studies in healthy people are brain scans. In 24 volunteers, 14 given Semax and 10 given placebo, a frontal brain network looked larger on resting MRI 5 and 20 minutes later (Lebedeva, 2018). In 52 volunteers given Semax, Selank or placebo, both drugs changed how the amygdala connected with the temporal lobe (Panikratova, 2020). Neither study gave a memory, attention or work test, so they show the drugs change brain activity, not that people think better (tested in people).
Patients Healthy volunteers. Healthy volunteers had brain scans only. Tested in people; none had a placebo group except the brain-scan studies.
No study has given Selank and Semax together, so nobody has measured what the combination does or whether it is safe.
The 2020 brain-scan study is the only one that used both drugs, and it gave each to a separate group. Online the two are often paired as calm plus focus. Their Russian labels conflict on one point: Semax lists anxiety disorders as a contraindication, while Selank is registered for them.
Neither is FDA-approved, and as of October 2026 neither can be legally compounded by a US pharmacy. Semax is closer to a change: an FDA advisory committee voted in its favor in July 2026, but FDA has not acted.
Every peptide's FDA, compounding and WADA status in one place: are peptides legal in the US.
These are the questions the published studies of both drugs leave open.
One email when a peptide trial reads out, a label changes or the FDA acts. Daily, free, every source linked.