Selank: the monograph

Selank is a synthetic seven-amino-acid peptide, sequence Thr-Lys-Pro-Arg-Pro-Gly-Pro, built by adding the tripeptide Pro-Gly-Pro to the immune peptide tuftsin¹²³. It was developed in Russia, where a 2022 peer-reviewed review of Russian peptide pharmaceuticals lists it among 14 peptide products registered in the Russian Federation and classifies it as an anxiolytic⁸. In the United States it is not an approved medicine of any kind: FDA's drug database returns no selank product, and FDA's compounding page carries a safety entry for selank acetate⁵⁶.

That gap is the whole story of this compound, and it is what this site is built to map. The same molecule is a registered pharmaceutical in one country and an unregulated research chemical in another, and the reason is not a conspiracy or a delay. It is that the evidence supporting selank was produced almost entirely inside one country's research system and has never been reproduced outside it. Our census of every selank record indexed in PubMed found 57 papers with selank in the title, 26 of them published in Russian, and exactly one carrying an author affiliation with no Russian institution, which is a US review article rather than an original selank study¹⁰¹¹.

This monograph does something no vendor page does and no critic page bothers with: it reads that literature at its real grade instead of laundering it into "clinically proven" or dismissing it as nonexistent. There are real human studies. They are small, they are Russian-language, their abstracts state no dose and no blinding, and nobody outside Russia has ever repeated them.

Every chip below carries an evidence grade. Human trial means a study in people, which for selank always means a small Russian study whose abstract omits dose and blinding. Animal means rodent, monkey or other non-human work. Regulatory means an FDA, WADA or register record. No chip on this page is supported by a randomized controlled efficacy trial, because no such trial of selank could be located¹⁰¹⁹.

Almost every page written about selank says it is an approved anxiolytic in Russia. We went looking for something citable behind that sentence, and here is exactly what we found and did not find.

What we can cite: a 2022 review of peptide biopharmaceutical development published in Pharmaceutics, written by authors at Russian Academy of Sciences institutes and Sechenov University, states that "14 products are registered in the Russian Federation" and gives the list. Selank is row 13, classified as an anxiolytic; semax is row 3, classified as a nootropic drug⁸. That is a peer-reviewed, fetch-verified statement by Russian pharmaceutical scientists about their own country's register.

What we could not do: verify it in the register itself. The Russian State Register of Medicines search requires a session-bound form submission, and its public autocomplete endpoint returned an empty result for selank. We did not treat that as evidence, because the control test failed: the same endpoint returned an empty result for paracetamol and aspirin, which are certainly registered⁹. An endpoint that cannot find aspirin cannot tell us anything about selank. So our statement is scoped to what the evidence actually supports: a peer-reviewed source says selank is a registered Russian peptide drug, and we could not independently confirm it at the register.

Registration in Russia is also not the same claim as efficacy by FDA or EMA standards, and it is worth being clear about why that is not a snub. Approval standards differ in what they demand: the trials described in the next section would not, on their published abstracts alone, support a US or European approval, because those abstracts do not report randomization, blinding, duration or dose¹³¹⁴¹⁵.

This is the section that exists nowhere else, so here is the method before the numbers. On 2026-08-14 we asked PubMed for every record matching selank and received 135. We fetched all 135 abstracts, kept the 57 with selank in the title, and counted two things per record: the publication language flag, and the country named inside the author affiliation block only, so that the word Russian appearing in an abstract could not be miscounted as an affiliation¹⁰. The script and its output ship in our verification folder, so the count can be rerun against the same source.

The results: 57 selank-titled papers. 26 of them are published in Russian¹⁰, in journals such as Zhurnal Nevrologii i Psikhiatrii imeni S.S. Korsakova, Eksperimental'naia i Klinicheskaia Farmakologiia and Bioorganicheskaia Khimiia. Of the 36 records that carry an author-affiliation block, 35 name a Russian institution. Exactly one does not, and it is a review article from the University of Connecticut and Hartford Hospital rather than original selank research¹⁰¹¹. The remaining 21 titled records predate PubMed's affiliation capture and are all in Russian-language journals.

Read that carefully, because the number that matters is a zero: the count of original selank research papers in the PubMed index from a non-Russian institution¹⁰. Thirty years of research on this compound, and no laboratory outside one country's research system has published an original study of it. Nobody has failed to replicate selank. Nobody has tried.

The clearest illustration is a 2020 paper from Sciensano, the Belgian federal government's scientific institute. Analysing two seized preparations, its authors described selank and semax as "the assumedly cognitive enhancing research peptides Selank and Semax, which, to our knowledge, have not completed any clinical trials"¹². Four Russian patient studies were sitting in PubMed when that sentence was published^13141516. They were invisible to a specialist government laboratory working directly on the compound, because they are in Russian, in journals a Western pharmacologist has no reason to read. That is not a claim that the Belgian authors were careless. It is what a language barrier looks like from the inside.

Two mechanisms are proposed, and they sit at different evidence grades. Neither has been confirmed in a human clinical study.

The first is enkephalin protection. Selank inhibits the enzymes that break down enkephalins, the body's own opioid peptides. In human serum it did so with a half-maximal inhibitory concentration of 20 micromolar, less potent than semax at 10 micromolar but far more potent than the reference inhibitors puromycin and bacitracin². A tritium-labelled assay in human plasma found selank is comparatively specific for carboxypeptidases and dicarboxypeptidases²⁴. In mice, a 100 microgram per kilogram dose produced an anxiolytic effect and lengthened plasma leu-enkephalin half-life in the anxious BALB/c strain, with no effect in C57BL/6 mice²⁵, and in patients with anxiety disorders shortened enkephalin half-life was itself an observed feature¹⁸.

The second is GABA modulation. A radioligand study reported that selank acts as a positive allosteric modulator of GABA binding and can block the modulatory activity of diazepam and olanzapine, with binding sites that "apparently not the same, but potentially may partially overlaps"²⁰. In rat frontal cortex a single 300 microgram per kilogram dose changed expression of 45 of 84 neurotransmission genes at one hour, with changes correlating with those produced by GABA itself²¹. In rat hippocampal slices selank increased spontaneous inhibitory postsynaptic currents in CA1 neurons, with no significant dose-dependence over the 1 to 8 micromolar range tested²².

One result cuts against the tidy version, and it belongs here rather than in a footnote. In cultured human IMR-32 neuroblastoma cells, selank on its own produced no change in any of the 84 GABAergic genes studied; it only altered what GABA and olanzapine did²³. A mechanism that shows up in rat cortex and disappears in human cells is exactly the kind of finding that needs independent work to resolve, and independent work is what this compound has never had.

Five human studies exist in the retrievable index. That is the entire human record, and it is short enough to read in full.

Zozulia 2008, 62 patients. The largest clinical comparison. Thirty patients with generalized anxiety disorder or neurasthenia received selank and 32 received the benzodiazepine medazepam. The reported result: "The anxiolytic effects of both drugs were similar but selank had also antiasthenic and psychostimulant effects"¹³. The same paper reported that patients had a shortened leu-enkephalin half-life correlating with symptom severity, which rose during selank treatment¹³.

Medvedev 2014, 60 patients. Selank against phenazepam in phobic-anxiety and somatoform disorders, reporting "pronounced anxiolytic and mild nootropic effects" and an anxiolytic effect that "lasted for a week after last receiving the peptide"¹⁴.

Medvedev 2015, 70 patients. Phenazepam alone in 30 patients against phenazepam plus selank in 40. The interesting result is not about anxiety but about side effects: adding selank "decreased the level of undesirable side-effects of phenazepam (attention and memory impairment, asthenia, sedation, increase in sleep duration, sexual disturbances, emotional indifference and orthostatism)"¹⁵.

Uchakina 2008, count not stated. Cytokine changes in serum of patients with generalized anxiety disorder and neurasthenia who received selank for 14 days, the only human study that states a duration¹⁶.

Panikratova 2020, 52 healthy participants. The only placebo-controlled study we located. Resting-state fMRI before and after injection of selank, semax or placebo, reporting differences in connectivity between the right amygdala and right temporal cortex, defined "for the first time"¹⁷. Note what the endpoint is: brain connectivity in healthy volunteers, not anxiety symptoms in patients.

Now the arithmetic and the caveats. Across every study with a stated participant count the total is 244 people: 192 patients plus 52 healthy volunteers, with the largest single study at 70¹⁰. None of the five abstracts states a dose. None states a route, except that Panikratova describes an injection. None reports randomization or blinding except Panikratova's placebo arm^1314151617. Four of the five are in Russian, so the full texts, which may well contain the design detail their abstracts omit, are not readable in English.

And no selank trial is registered anywhere: an intervention-scoped ClinicalTrials.gov query returns two records, both unrelated, matched only because selank's synonym TP-7 is also a scalp electrode position¹⁹. No randomized controlled efficacy trial of selank was located in any language index we can read.

The bulk of the selank literature is rodent work, and it is where nearly every benefit sold on a vendor page actually originates. The full graded table sits below; the pattern is worth stating first.

Anxiety and stress: in rats under unpredictable chronic mild stress, selank alone most effectively reduced elevated anxiety, and the diazepam plus selank combination was most effective under chronic stress²⁶. Alcohol and opioids are the best-developed thread: selank at 0.3 milligrams per kilogram daily prevented ethanol-induced memory and attention disturbance during withdrawal and blocked the ethanol-induced BDNF rise²⁷; the same dose reduced the total morphine withdrawal index by 39.6 percent against diazepam's 49.3 percent²⁸; and in DBA/2 mice it prevented ethanol-induced hyperlocomotion²⁹. Stress physiology beyond behaviour: a Kursk group reported that 80, 250 and 750 microgram per kilogram doses reduced corticosterone and the pathomorphological changes of restraint stress in rat colon, and restored stress-disrupted gut microbiota³⁴³⁵. Immune effects are real and measurable: a single 100 microgram per kilogram dose changed expression of 34 of 84 inflammation genes in mouse spleen³, and selank suppressed influenza replication in cell culture and in mice, most effectively when given before inoculation³⁶. In monkeys, intranasal selank reduced fear and aggression in experimental neurosis³².

The negative results belong in the same paragraph as the positive ones. In rats with 6-hydroxydopamine-induced parkinsonism, neither selank nor semax affected motor activity or passive defensive behaviour³⁷. Reporting that is not a hedge; a literature where everything works is a literature to be suspicious of.

The species point is the one to carry away. A 39.6 percent reduction in a rat withdrawal index is a real finding about rats. It is not a dose you can scale to a person, because the human pharmacokinetics needed to do that translation have never been measured¹⁰³³.

Every row below carries the species or model it was performed in, the dose and route as the source states them, and the outcome as reported. Where an abstract does not state a group size, the row says so rather than inventing one. The census: 6 human rows, 15 animal rows and 5 in vitro rows. Not one row is a randomized controlled efficacy trial¹⁰.

There is no human pharmacokinetic study of selank. No half-life, no peak concentration, no clearance figure and no bioavailability value for any route has been published in the index we can read¹⁰. A term scan of all 135 fetched records found no such study.

What exists is one Russian tritium-labelling study, which identified the pentapeptide TKPRP, the tripeptide TKP and the dipeptides RP and GP as the major products of selank breakdown in blood plasma, and examined selank in rat brain tissue after intranasal dosing. Its abstract reports no numeric parameters³³.

One correction worth making explicitly, because it circulates widely. Half-life figures quoted for selank are almost always the half-life of leu-enkephalin, the substrate whose breakdown selank slows, and not of selank itself. The two studies that mention a half-life both measure enkephalin¹⁸²⁵. If you see a confident number for selank's half-life, ask which molecule it refers to.

Selank is sold in the United States in two forms, both documented by the Belgian government laboratory that analysed seized preparations: "freely available either as lyophilized powder for injection purposes or are present in nasal sprays"¹². The nasal form is the one most commonly encountered.

The research used intranasal and intraperitoneal routes, not the subcutaneous injection familiar from other peptide markets. Intranasal dosing appears in mice, rats and monkeys^30313233; intraperitoneal dosing dominates the rodent behavioural work. No fetched study used subcutaneous injection¹⁰.

The most useful study for anyone comparing forms is a direct head-to-head in mice, and its result is not the one either marketing side expects. At 300 micrograms per kilogram per day for five days, both routes produced anxiolytic and nootropic effects, but only in the anxious BALB/c strain, and the two routes acted on different receptor systems: intraperitoneal dosing raised GABA-receptor binding sites in frontal cortex by 38 percent with no NMDA change, while intranasal dosing raised NMDA-receptor binding sites by 23 percent with no GABA change³⁰. The authors attributed the difference to pharmacokinetics and biotransformation. In mice, the route changes what the drug does, not merely how much arrives. Whether that holds in humans is unknown, because no human study of either route reports a dose at all¹³¹⁷.

No. There is no established human dose of selank, and the reason is specific rather than rhetorical: not one of the five human studies states a dose in its abstract^13141517. Four are in Russian and their full texts were not retrievable in this run, so a dose may well be printed inside them; we simply cannot cite what we have not read.

What can be cited are the animal regimens, and they are worth seeing because of how the arithmetic works. The rodent behavioural studies cluster at 0.3 milligrams per kilogram²⁷²⁸²⁹, with gene-expression work at 100 to 300 micrograms per kilogram²¹³, the stress-physiology studies at 80 to 750 micrograms per kilogram³⁴, and the route comparison at 300 micrograms per kilogram³⁰. Our study-dose explorer lists each of these with its species attached.

What we will not do is convert them. Scaling a rat intraperitoneal dose to a human intranasal dose requires knowing human bioavailability by that route, and that number does not exist for selank¹⁰. Any milligram figure you see presented as a selank protocol was not derived from a published human study. This site therefore ships no dose calculator, and the reason is printed on the tool page itself.

The honest answer is that nobody knows, and the specific gaps matter more than the general uncertainty.

What is reported: the Russian clinical studies describe selank as well tolerated, with one assessing tolerability on the UKU side-effect scale¹⁴ and another finding that adding selank reduced the side effects of a benzodiazepine¹⁵. Those are the reassuring data points, and they come from short courses in small groups.

What does not exist: any long-term safety study, any formal toxicology report, and any adverse-event surveillance for selank in the index we can read¹⁰. Only one human study even states how long people took it, and that was 14 days¹⁶.

The gap FDA names is immunogenicity, and it is a pointed one for this molecule. Selank is a tuftsin analog, and tuftsin is a fragment of the immunoglobulin G heavy chain, so this is a peptide with an immune pedigree being given to people with no published assessment of whether the immune system reacts to it⁶³. FDA's stated concern is immunogenicity risk from aggregation and peptide-related impurities, and that it lacks important information about any safety issues raised by selank administered to humans⁶. The animal literature does contain immune pharmacology, including spleen inflammation-gene expression and antiviral activity³³⁶, but that is a different question from whether the body mounts a response against the peptide itself.

This question cannot be answered from evidence, because the studies that would answer it were never done. What exists is a list of groups with no data at all: pregnant and breastfeeding people, children and adolescents, older adults, people with liver or kidney impairment, people with autoimmune conditions or on immunosuppressants (relevant given the tuftsin lineage and the reported immune effects³³⁶), and anyone with a psychiatric diagnosis being managed with prescribed medication.

That list is not a set of established risks. It is a set of populations in which selank has never been studied, which is a different and in some ways less comfortable statement¹⁰.

No formal drug-interaction study of selank exists. Three findings are nonetheless directly relevant, and two of them involve benzodiazepines.

The human one: in 70 patients, adding selank to phenazepam reduced that benzodiazepine's side effects, including attention and memory impairment, sedation and emotional indifference¹⁵. The rat one: under unpredictable chronic mild stress, the diazepam plus selank combination was the most effective condition tested²⁶. The mechanistic one cuts the other way: selank blocked the modulatory activity of diazepam and olanzapine at the GABA receptor in a binding study²⁰, and in human IMR-32 cells selank suppressed the gene-expression changes GABA produced alone while amplifying those of olanzapine²³.

A compound that interacts measurably with benzodiazepines and antipsychotics in every system it has been tested in, and that has no human interaction study, is not a compound to combine with prescribed psychiatric medication on the basis of a forum protocol.

Nobody is checking. There is no pharmacopoeial monograph for selank, no validated quality standard, and no regulator verifying the contents of a research-market vial or nasal spray sold in the United States. FDA states that even for compounded drugs, which are a more regulated category than research chemicals, it "does not verify the safety, effectiveness or quality" before marketing⁷.

The scale of the identification problem is visible in how the Belgian government laboratory had to proceed. To identify what was in the seized preparations, its scientists had to develop a new liquid chromatography tandem mass spectrometry method, then validate it across matrices to ISO 17025, because standard screening was not equipped for these peptides¹². If a national control laboratory needed a bespoke analytical method, a certificate of analysis supplied by the seller is not an independent check.

The regulatory description of the US market is blunt in the peer-reviewed literature too. A 2021 review in the Journal of Clinical Pharmacology describes phenibut and selank as "poorly studied Russian drugs with GABAergic mechanisms that are inexplicably sold to US consumers as dietary supplements"¹¹.

Selank is not named anywhere on the 2026 WADA Prohibited List³⁸. That is not the same as being permitted, and the distinction is where athletes get caught out.

The S0 category prohibits at all times "any pharmacological substance which is not addressed by any of the subsequent sections of the List and with no current approval by any governmental regulatory health authority for human therapeutic use"³⁸. Whether selank falls inside that catch-all depends on whether its Russian registration counts as current approval by a governmental regulatory health authority. We found no source resolving that question for selank by name³⁸³⁹.

The anti-doping community has not resolved it either. A 2025 review co-authored by the Polish Anti-Doping Agency's scientific team and the anti-doping research centre at the University of Lausanne places selank in its table of substances with unclear status. Semax, in the row directly below, is flagged as similar to a prohibited substance; selank's equivalent column is empty³⁹. The same paper warns that the open-ended nature of the prohibited list "may lead to unintentional violations of anti-doping rules"³⁹.

For a tested athlete the practical answer is simple even though the classification is not: unclear status is not clearance, and the only safe route is a written determination from your own anti-doping organisation before use, not an inference from the list's silence.

Two comparisons are worth making with sources attached. The first is selank against semax, the peptide it is most often confused with and most often stacked with. The second is the same molecule seen from two regulatory systems, which is the clearest way to show what a research-chemical status actually means. We run a sister reference site on semax at semaxlabs.com, and the rows below cite primary sources rather than either site's opinion.

See the evidence-limits block. It is the honest core of this monograph, and on this compound it is longer than the list of things we do know.

Twelve questions, each answered in 320 characters or fewer, every answer consistent with the graded evidence above.

Every reference below was fetched and verified on 2026-08-14: PubMed records through the NCBI E-utilities API, open full texts through Europe PMC, FDA pages retrieved directly from fda.gov with a browser user agent, chemical identity from PubChem, trial absence from the ClinicalTrials.gov v2 API, and the WADA list from an archived same-day PDF. Russian-language titles are marked as such. A machine-readable version of this list ships as the citation manifest for this page.

Semax (semaxlabs.com) is the closest relative: another Russian heptapeptide from the same research programme, an ACTH(4-10) analog classified as a nootropic where selank is classified as an anxiolytic⁸⁴⁰. The two appear together throughout this literature, share the enkephalinase mechanism², and were found together in the same seized preparations¹². They also have the same underlying evidence problem, and reading either site should leave you with the same caution.