Selank vs Semax: two Soviet-era peptides, two cognitive angles

You go looking for something that sharpens focus and quiets the noise. You hit a forum thread comparing Selank and Semax. Same lab, same era, same Cyrillic-only literature trail, often sold next to each other as cognitive peptides. The thread tells you to pick.

The thread is asking the wrong question. The two peptides were built for two different conditions, the human evidence sits in two different places, and the FDA is treating them differently. Stack them up against each other on focus and you are choosing between two compounds neither of which was studied for focus in the first place.

Here is what’s actually under the hood for each one, and what splits them apart.

Where they both came from

Both peptides come out of one institute and one logic.

The Russian Academy of Medical Sciences group around Nikolai Miasoedov spent the late Soviet period working from one shared idea. The body already makes neuroactive peptides. They are usually broken down too fast to use as drugs. What if you took a fragment of one of them and stabilised it.

Selank is the answer when the starting molecule is tuftsin — a four-amino-acid fragment of an immune-system protein with mild anxiolytic and immune-modulating activity. Add a three-amino-acid Pro-Gly-Pro tail to slow it down enzymatically and selank is what you get: a seven-residue peptide. The pitch was a peptide anxiolytic without the sedation, dependence, and cognitive blunt of a benzodiazepine.

Semax is the answer when the starting molecule is ACTH, the pituitary stress hormone whose N-terminal fragments carry most of the molecule’s neuroactivity. Take the ACTH(4-7) core — Met-Glu-His-Phe — and add the same Pro-Gly-Pro tail, and semax is the seven-residue peptide you end up with. Western pharmacology had separately spent the 1970s and 1980s studying ACTH(4-10), a related but different fragment, for attention and memory, and never quite landed a clinical claim. The Miasoedov group’s stabilised version — same MEHF head, PGP tail instead of the rest of ACTH — was developed for acute ischemic stroke and registered in Russia for that.

So one is an anxiolytic-class peptide. The other is a neuroprotective-class peptide. The cognitive-performance angle the western internet sells them on is not the angle either was built around.

What the selank human file actually shows

This is the part that surprises people. Selank has more controlled human data than most of the peptides you read about online — it is just that all of it is in Russian.

Two trials are load-bearing.

The 2008 Zozulia trial ran 62 patients with generalised anxiety disorder and neurasthenia. Half received selank, half received medazepam — a benzodiazepine with a long Russian and European prescribing history. The anxiolytic effects were similar across both arms. What was different is what came alongside the anxiolysis. Selank also produced antiasthenic and mild psychostimulant effects, the opposite of the cognitive drag a benzodiazepine drops on you. The investigators tracked an enkephalin biomarker in blood serum as a mechanistic correlate, and it moved in the direction the theory predicted.

The 2014 Medvedev trial ran 60 patients with phobic-anxiety and somatoform disorders under ICD-10 codes F40 to F45. Selank was compared against phenazepam, another benzodiazepine. The result was a pronounced anxiolytic effect plus a mild nootropic effect, and a quietly surprising tail. The anxiolytic action lasted around a week after the patient took their last dose, which is not what benzodiazepines do.

Treat that picture honestly. The Russian register has selank as a registered anxiolytic, two controlled trials show comparable anxiolysis to standard-of-care benzodiazepines, and the residual nootropic and antiasthenic effects are real signals worth taking seriously. The picture also has limits: roughly 120 patients of combined evidence, single research cluster, no peer-reviewed western replication, and no work studying selank in the population the online market is actually selling to — healthy adults chasing focus, not GAD patients seeking treatment.

What the semax human file actually shows

Semax’s evidence base sits in stroke neurology, not productivity.

The 1997 Gusev trial is the canonical one. Thirty patients in the acute period of hemispheric ischemic stroke received semax on top of conventional therapy; eighty patients with comparable stroke severity formed the control group. The investigators tracked clinical rating scales, EEG, and somatosensory evoked potentials. Adding semax accelerated the regress of general cerebral symptoms and, in particular, motor neurological deficits. The trial was the foundation for the Russian Federation’s eventual registration of semax as a cerebroprotective agent for ischemic stroke.

Dolotov and colleagues, working in 2006, gave the mechanism a clean read. Semax binds specifically in rat-brain tissue and increases brain-derived neurotrophic factor — BDNF, the growth factor most commonly associated with neuronal survival and synaptic plasticity. If a peptide is going to do anything for a brain recovering from oxygen starvation, BDNF is roughly the thing you want it to push on, and the mechanism work supports the idea that semax does.

Honest read on the semax file: the data is real, narrow, and sits squarely in acute stroke. There is downstream Russian work on semax in other neurological settings — chronic cerebral ischemia, migraine, attention disorders in children — and it is also single-cluster Russian literature. The cognitive-performance pitch in the supplement scene is extrapolated from the stroke and BDNF data, not from any controlled trial in healthy adults asking the cognition question directly.

The Russian-literature problem

Both peptides share the same evidence-base weakness, and it is worth saying out loud.

Korsakov’s Journal of Neurology and Psychiatry — the publication that carries most of this work — is a legitimate peer-reviewed Russian psychiatric journal. The trials cited above were published there, are indexed in PubMed, and have abstracts in English. But the underlying methods, statistical analysis, and follow-up sections often live only in the Russian text, and the replication arc that would normally follow a positive controlled trial — independent labs in other countries running similar studies and publishing in English-language journals — has not happened for either peptide.

The 2026 Sports Medicine review by Mendias and Awan is candid about the broader injectable-peptide pattern: a parallel grey market of unapproved compounds operates largely outside regulatory oversight, with rigorous human safety data scarce and potential for serious patient harm. The Mendias review does not name selank or semax in its twelve-peptide list, which is itself the point — the Russian Academy peptide tradition sits outside the Western injectable-peptide review literature even as it sits inside the same broader adoption-ahead-of-evidence dynamic. There is real human data on selank and semax, conducted seriously, in real populations, by a real institute. There is also a closed evidence loop, with no outside scrutiny telling you whether the findings reproduce when somebody who doesn’t already believe in the peptide tries to find them.

So what’s actually different between them

Strip the marketing back. The honest split is this:

Selank is the anxiety-and-asthenia peptide. The controlled human evidence supports anxiolytic effects with a side of cognitive non-impairment that’s the opposite of what benzodiazepines do. If you are a healthy adult chasing focus, the population you are most like is not the population selank was studied in.

Semax is the acute-cerebral-injury peptide. The controlled human evidence supports faster recovery of motor function after ischemic stroke, and a BDNF-elevating mechanism that’s at least directionally compatible with cognitive applications. If you are a healthy adult chasing focus, you are also not the population semax was studied in.

The implication is uncomfortable but clean. Picking selank or semax for focus is not a choice between two compounds with comparable evidence for focus. It is a choice between two compounds with evidence in two completely different clinical settings, neither of which is your setting.

Where regulators stand — and where the two peptides split

This is the place where the two stop looking similar at all.

Semax is on the July 23–24, 2026 FDA Pharmacy Compounding Advisory Committee docket per the April 2026 Federal Register notice. The proposed indications listed for semax are cerebral ischemia, migraine, and trigeminal neuralgia — neurology indications, not the cognitive-performance ones the consumer market is buying it for. If PCAC says yes, US licensed pharmacies will be able to compound semax for a patient with a prescription for one of those conditions, under Section 503A. That is meaningfully more legitimate than today’s grey-market path. It is not blanket approval, and it isn’t approval for focus or longevity uses.

Selank is not on the docket at all. The same Federal Register notice that establishes the meeting and lists the nominated peptides does not include selank. That leaves selank in the same regulatory grey zone as a year ago — no FDA approval, no compounding pathway in motion, sold through the same channels that sell everything else in this corner of the market. If you are buying selank in 2026, you are still buying research-chemical-grade material from an unregulated supply.

For competitive athletes, the WADA picture is currently the same for both — neither is on the prohibited list. As always with WADA, check on the day of testing, not from this article.

What would actually settle this

If you want to make a confident decision about either peptide, here is the short list of evidence neither one has yet.

Independent Western replication of at least one Russian trial per peptide, in a peer-reviewed English-language journal, by a group that doesn’t share authorship with the Miasoedov institute. That is the move that closes the credibility gap.

A controlled trial of selank in the actual population people are buying it for — healthy adults reporting attention or anxiety symptoms below the clinical threshold for diagnosis. Whether the GAD signal extrapolates down to subclinical anxiety is not something the existing trials answer.

A controlled trial of semax in healthy adults asking the cognitive question directly — working memory, executive function, attention — instead of stroke recovery. The mechanism makes this plausible. The data does not yet exist.

And, for both, real human pharmacokinetic and long-term safety data. Russian regulatory registration is not the same as a full FDA development programme, and the safety profile in a 40-year-old who plans to dose for five years is not the same question as the safety profile in a stroke patient on a ten-day course.

So where does that leave you

If you came here to pick the better cognitive peptide between selank and semax, the honest answer is that the question doesn’t have a clean answer with the evidence that exists. Both have more human data than most peptides in this scene, both have it in narrow clinical populations, and neither has it in you.

If you came here to figure out whether the marketing is doing you a service, it isn’t. The vs-page framing online flattens two compounds with different mechanisms, different evidence bases, and different regulatory trajectories into a head-to-head that doesn’t reflect what either one was tested against.

The honest move is to wait. Wait for the evidence to come in. Wait for an independent lab to replicate. Wait, if you are American, for what the PCAC review does for semax in July. And in the meantime, take seriously that no good answer yet is itself an answer.

That is what we are building. Wolverine Health is a physician-supervised peptide service — real prescriptions, US-licensed compounding pharmacies, every batch third-party tested. It isn’t live yet, because the regulation isn’t either, and we’re not going to sell you ahead of the science. If you want to know the moment Semax actually comes out of PCAC in July, or the moment Selank gets a Western RCT that closes the replication gap — with the same honesty you just read — join the waitlist.