Semax Dosage: Chart, Intranasal vs SubQ - Dosage Peptide

Semax Dosage: Chart, Intranasal vs SubQ - Dosage Peptide Semax dosage explained: the 0.1% and 1% solution math, 5mg/10mg reconstitution charts, intranasal versus subQ, and the real evidence tier behind each claim. Semax (Met-Glu-His-Phe-Pro-Gly-Pro) is one of the few research peptides where the dosing question has a real, documented anchor — and also one where that anchor is routinely misrepresented. Anyone searching for a semax dosage figure is usually holding a lyophilised 5 mg or 10 mg vial, while every published human number comes from a registered Russian intranasal solution described as a percentage, not from a vial and a syringe. This article reconstructs the arithmetic that connects those two worlds, states precisely what tier of evidence supports each number, and is explicit about where the published record simply runs out — most importantly for the subcutaneous route, which has almost no human dosing literature at all. Semax is a synthetic heptapeptide with the sequence Met-Glu-His-Phe-Pro-Gly-Pro (single-letter: MEHFPGP), molecular formula C37H51N9O10S and a molecular weight of approximately 813.9 Da[1]. Structurally it is the ACTH(4-7) fragment of adrenocorticotropic hormone — Met-Glu-His-Phe — extended at the C-terminus with the tripeptide Pro-Gly-Pro. It is therefore usually described in the literature as an analogue of the ACTH(4-10) fragment, and the two descriptions (ACTH(4-7)-PGP and “ACTH(4-10) analogue”) refer to the same molecule. The registered Russian product labelling uses the ACTH(4-10) formulation[2]. The compound emerged from Soviet and then Russian neuropeptide research programmes, principally at the Institute of Molecular Genetics of the Russian Academy of Sciences (now within the National Research Centre “Kurchatov Institute”), and the name of N. F. Myasoedov appears on a large fraction of the primary literature spanning three decades. That single-lineage provenance matters for how the evidence should be weighted, and it is addressed directly in the sections on evidence level and limitations below rather than left implicit. The reason the molecule is not simply ACTH(4-7) is pharmacokinetic. Short native peptide fragments are degraded rapidly by aminopeptidases and other proteases, which limits how long an unmodified fragment can act. Reference[4] characterises Semax in its opening line as an ACTH(4-10) fragment analogue with prolonged neurotropic activity. The attribution of that prolongation specifically to the C-terminal Pro-Gly-Pro extension — the standard mechanistic account repeated throughout the field — is not established by the source cited here, and this article will not pretend otherwise. It is a widely held, chemically reasonable account (proline-rich C-termini are a well-known general strategy for resisting exopeptidase cleavage), but the reader should file it as the field’s working explanation rather than as a demonstrated result. Two related caveats belong here, because they are where secondary Semax content characteristically inflates. First, figures such as “a 20–24 hour half-life” circulate widely on vendor and aggregator pages. That specific number is not something this article will assert, because it could not be traced to a primary source stating it. What the accessible literature supports is the qualitative claim — prolonged neurotropic activity relative to the bare fragment — not a precise half-life figure. Where a number cannot be sourced, this article says so rather than repeating it. Second, the published stability work on Semax that is indexed in the international literature concerns proteolytic stability in biological media — how the peptide resists enzymatic degradation in serum, plasma or tissue homogenate — and, in the most frequently cited case, examines an acetylated derivative rather than unmodified Semax. That is a different question from the one a researcher holding a vial actually has, which is chemical stability in bacteriostatic water in a refrigerator over weeks. The two are routinely conflated. They should not be. Because Semax is derived from ACTH, a reasonable first assumption is that it stimulates the adrenal cortex and raises cortisol. The literature consistently says it does not. The ACTH(4-10) region carries the behavioural and attentional activity of adrenocorticotropin without the steroidogenic activity, which resides in a different portion of the parent hormone. The evidence tier here needs stating carefully, because this claim is usually presented with more force than its sources carry. The registered Russian product labelling describes Semax as a synthetic ACTH(4-10) analogue “completely devoid of hormonal activity”[2], and a 2024 study in the European Journal of Pharmacology classifies both Semax and Melanotan II as “noncorticotropic synthetic analogs of the ACTH(4-10) fragment”[5]. Both are descriptive characterisations rather than measurements: no study in this reference set assayed cortisol or adrenal steroidogenesis after Semax administration. The claim rests on the pharmacology of the parent hormone — steroidogenic activity resides outside the 4-10 region — and on consistent, unchallenged description in the primary literature and in an approved product label. That is a reasonable position to hold, but it is held by convention and by structural inference, not by a direct experiment cited here. It is worth stating precisely all the same, because it is frequently garbled in both directions: into “Semax is a corticosteroid” or into “Semax boosts ACTH”. Neither is what the literature says. Readers approaching Semax through its relationship to the melanocortin system may also want the adjacent comparison work on how Adamax and other ACTH-derived neuropeptides differ from Semax in the research literature, and the deeper treatment of how Semax influences ACTH-related pathways in experimental models. This is the section that determines how every dosing number below should be read, so it is placed before the numbers rather than after them. Semax is not FDA-approved. There is no US marketing authorisation for Semax for any indication. It is not an approved drug and it is not a lawful dietary supplement ingredient. Semax is not EMA-approved. There is no European centralised marketing authorisation, and no EMA assessment report exists to consult. Semax is a registered medicine in Russia (and has been used in some post-Soviet states). Two strengths are marketed as nasal drops: a 0.1% solution[2] and a 1% solution[3], both manufactured by Peptogen INPC, both prescription-only, and both supplied in 3 mL dropper bottles. The primary literature refers to this clinical use directly and routinely — describing Semax as a peptide that “is successfully used for acute stroke therapy”[6] and as “a synthetic melanocortin derivative that is used in the treatment of ischemic stroke”[7]. Everything sold as a lyophilised 5 mg or 10 mg “Semax” vial in the Western research-chemical market is not the registered product. It is a research chemical of independently variable purity, identity, and sterility, in a presentation the registered product does not use. The honest one-sentence version: Semax is registered in Russia, not FDA- or EMA-approved; the clinical evidence base is largely Russian-language, frequently small, and has not been independently replicated to Western regulatory standards. Because the entire dosing discussion below is built on the registered product, it is worth setting out what that product’s labelling says rather than paraphrasing it loosely. The two strengths carry different indication sets, and conflating them is a common source of error. The 0.1% strength is labelled for intellectual-mnestic disorders arising from cerebrovascular disease, recovery after stroke, dyscirculatory encephalopathy, transient ischaemic attacks, recovery following head trauma, neurosurgery and anaesthesia, neurotic disorders, adaptation under extreme conditions, prevention of mental fatigue, optic nerve atrophy and optic neuritis, and minimal brain dysfunction in children aged seven and older[2]. The 1% strength carries a single, much narrower registered indication: ischaemic stroke in the acute period[3]. Two things follow that are worth holding onto. First, this is a broad indication list by Western regulatory standards — “adaptation under extreme conditions” and “prevention of mental fatigue” are not indications any FDA- or EMA-approved neurological drug carries, and their presence on a label is a fact about the registering system as much as about the molecule. A registered indication is a regulatory decision made under a particular evidentiary standard; it is not itself evidence of efficacy, and it does not transfer across jurisdictions. Second, the indication list is the closest thing that exists to an official statement of what Semax is for — and none of the entries is “cognitive enhancement in healthy adults”, which is what most people searching for a Semax dose actually have in mind. A search of ClinicalTrials.gov — the NIH registry of interventional studies — for Semax as an intervention returns zero registered trials[8]. Not zero completed trials; zero registered trials of any status, at the time of writing. Whatever one concludes about Semax, this is the single most informative data point about its evidentiary position: a compound registered as a medicine in one jurisdiction since the 1990s has generated no registered trial activity in the international registry that the FDA, EMA, and ICMJE journals treat as the baseline for trial transparency. This is not proof that Semax does not work. Registration on ClinicalTrials.gov is a regulatory and publishing convention, and a drug developed entirely inside a domestic Russian framework in the 1990s had no procedural reason to appear there. Registries are also a relatively recent institution: prospective registration only became an ICMJE publication condition in 2005, by which time the pivotal Semax work was already eight years old. But the consequence stands regardless of the reason. Anyone who claims Semax is “clinically proven” is claiming something that cannot be checked in the way clinical claims are normally checked — no protocol to compare against the published report, no way to detect selective outcome reporting, no denominator for how many studies were run. The mechanistic literature on Semax is genuinely substantial, and it is also almost entirely preclinical. This is worth separating cleanly from the clinical question. Rodent and cell-culture work can establish that a molecule engages a biological system; it cannot establish that engagement produces a clinical benefit in humans. The most-replicated mechanistic finding is that Semax raises brain-derived neurotrophic factor. A 2006 study in the Journal of Neurochemistry is the cleanest example: intranasal Semax at 50 and 250 µg/kg body weight in rats produced a rapid increase in BDNF protein in the basal forebrain at 3 hours, but not in the cerebellum — a regionally specific effect rather than a global one. The same study characterised specific, reversible, calcium-dependent binding of tritium-labelled Semax to basal forebrain cell membranes, with a dissociation constant KD of 2.4 ± 1.0 nM and a Bmax of 33.5 ± 7.9 fmol/mg protein[9]. A nanomolar KD is a real binding affinity, not an artefact, and it is the strongest single piece of mechanistic evidence in the Semax file. Downstream of the binding work, a study in a permanent middle cerebral artery occlusion (pMCAO) model examined transcription of neurotrophins and their receptors in rat cortex after treatment with either Semax or the PGP tripeptide alone. The results are more specific than they are usually reported. Semax enhanced transcription of Bdnf, TrkC and TrkA at 3 hours after occlusion, and of Nt-3 and Ngf at 24 hours, with Ngf still enhanced at 72 hours. PGP alone enhanced Bdnf and TrkC at 3 hours, and Ngf, TrkB, TrkC and TrkA at 24 hours[6]. The commonly repeated summary “Semax activates Bdnf, Ngf and TrkB after ischaemia” is wrong on the third gene. That trio comes from the same paper’s one-line description of earlier work in intact rat hippocampus, not from the pMCAO experiment; in the ischaemia model, TrkB was enhanced by PGP, not by Semax. This is a small error with a large lesson: the drift happens when a summary sentence in an abstract’s background is mistaken for the abstract’s result. That the PGP tripeptide alone reproduces part of the effect is worth noting — though the authors qualify it directly, reporting that Semax selectively affected transcription of neurotrophins and their receptors in the ischaemic rat cortex, whereas the influence of PGP was “mainly unspecific”[6]. The “tail” added for stability is not pharmacologically inert, but the authors do not claim it is doing the same job as the whole molecule, and neither should anyone citing them. The most methodologically modern strand is genome-wide. A 2014 BMC Genomics study performed genome-wide transcriptional analysis of Semax effects in ischaemic rat brain cortex at 3 and 24 hours after permanent MCAO, finding predominant effects on immune- and vascular-system gene expression[10]. In the transient MCAO (ischaemia-reperfusion) model, genome-wide RNA-Seq analysis reported that Semax suppressed inflammatory gene expression and activated neurotransmitter gene expression, and a 2021 International Journal of Molecular Sciences paper extended the picture to the protein level: at 24 hours after tMCAO, active CREB was upregulated in subcortical structures including the focus of ischaemic damage, MMP-9 and c-Fos were downregulated in the adjacent frontoparietal cortex, and active JNK was downregulated in both tissues under Semax[7]. That is a coherent mechanistic story — inflammation and cell-death signalling down, a recovery-associated transcription factor up — and it is entirely in rats, entirely in surgical ischaemia models, and entirely from one institutional lineage. Coherence is not replication. Chronic intranasal administration in white rats at 0.05 mg/kg daily for 10 or 14 days produced anxiolytic and antidepressant-like effects without altering exploratory activity in a non-stressogenic environment; the authors attributed these effects to activation of the brain serotonergic system together with increased hippocampal BDNF expression[15]. Note the structure of that claim: the behavioural finding is a result, the serotonergic-plus-BDNF explanation is the authors’ interpretation offered in the same breath. A 2024 chronic unpredictable stress study reported antidepressant-like and antistress effects in male rats given 60 nmol/kg intraperitoneally[5]. Both are animal models of affective behaviour, which are notoriously weak predictors of human antidepressant efficacy; the appropriate reading is “the molecule is behaviourally active in rodents”, not “Semax has antidepressant effects”. For a broader treatment of the circuit-level questions this raises, see the discussion of how Semax might affect neural circuit stability under cognitive load. The registered Russian product is a nasal solution described by percentage concentration. Converting that to mg/mL is straightforward arithmetic, and it is the bridge between the clinical literature and a vial of powder. Percentage concentration in pharmaceutical solutions conventionally means weight/volume: X% w/v = X grams per 100 mL. The labelling confirms this reading directly, stating 1 mg per 1 mL for the 0.1% strength[2] and 10 mg per 1 mL for the 1% strength[3]. 0.1% 0.1 g per 100 mL = 100 mg per 100 mL 1 mg/mL 1,000 mcg/mL 3 mg 1% 1 g per 100 mL = 1,000 mg per 100 mL 10 mg/mL 10,000 mcg/mL 30 mg So the 1% solution is exactly ten times the concentration of the 0.1% solution. This single conversion resolves most of the confusion in circulating Semax dosing content, where percentage figures and mg figures are mixed without reconciliation. It also makes the scale of the strengths concrete: an entire 3 mL bottle of the 0.1% product contains 3 mg of peptide — less than a third of what a single research-chemical 10 mg vial claims to hold. The “about 50 mcg per drop” figure that circulates on aggregator pages is, unusually, traceable — but only to the registered product’s own labelling, and only for the registered product’s own dropper. The labelling for the 0.1% solution states that one drop contains 50 mcg of active substance[2], and the labelling for the 1% solution states a drop volume of 0.05 mL containing 500 mcg[3]. Those two statements are internally consistent: 0.05 mL (50 µL) of a 1 mg/mL solution is 50 mcg, and 50 µL of a 10 mg/mL solution is 500 mcg. The drop is the same size; only the concentration changes. The important qualification is what that number is a property of. It is a specification of the dropper supplied with the registered product, not a property of Semax solutions in general. Droplet volume is determined by the geometry and surface characteristics of the dispensing tip, the viscosity and surface tension of the liquid, and the angle at which it is held. A reconstituted research vial fitted with an arbitrary dropper has no established drop volume, and transferring “50 mcg per drop” to it is an assumption dressed as a measurement. The percentage-to-mg/mL conversion above is arithmetic and holds universally. The per-drop figure holds for one specific manufactured device, and that distinction is the single most common place where Semax dosing content silently invents precision. Below is what published human studies and the registered labelling actually specify. This is a description of research protocols and approved-product labelling, not a recommendation, and every row carries its citation so the claim can be checked. Registered 1% labelling Acute ischaemic stroke (registered indication) 2–3 drops per nostril (moderate); 3–4 drops per nostril (severe). 1 drop = 0.05 mL = 500 mcg Intranasal 3–4 times daily (moderate); 4–5 times daily (severe); 10 days [3] Registered 0.1% labelling Cognitive/cerebrovascular indications, optic nerve pathology, others 2–3 drops per nostril per administration. 1 drop = 50 mcg Varies by indication [2] Gusev et al., 1997 30 patients, acute hemispheric ischaemic stroke (vs 80 conventional-therapy controls) 12 mg/day (moderate strokes); 18 mg/day (severe strokes) 5-day course (moderate); 10-day course (severe) [11] Gusev, Martynov, Kostenko et al., 2018 110 post-ischaemic-stroke patients in rehabilitation 6,000 mcg/day (= 6 mg/day) Not specified in abstract Two 10-day courses separated by a 20-day interval [12] Lebedeva et al., 2018 (resting-state fMRI) 24 healthy volunteers (14 Semax, 10 placebo), mean age 43.9 ± 9.5 1% solution (= 10 mg/mL); total dose not stated in abstract Single administration; imaging at baseline, 5 min, 20 min [13] Panikratova et al., 2020 (functional connectomics) 52 healthy participants (Semax, Selank, or placebo) Not stated in abstract “Injection” — subtype not specified in abstract Single administration; imaging before, 5 min, 20 min [14] One useful piece of internal validation is available here, and it is worth doing explicitly because it is checkable. The 1% labelling describes, for moderate strokes, 2–3 drops per nostril, 3–4 times daily[3]. Taking the upper end: 3 drops × 2 nostrils = 6 drops per administration; 6 × 500 mcg = 3,000 mcg per administration; 3,000 mcg × 4 administrations = 12,000 mcg = 12 mg/day. That is exactly the moderate-stroke figure reported in the 1997 clinical study[11]. For severe strokes the label’s 3–4 drops per nostril, 4–5 times daily spans roughly 12–20 mg/day, within which the study’s 18 mg/day sits comfortably. This tells us the labelling and the pivotal trial describe the same regimen, which is reassuring about the internal consistency of the Russian record. It tells us nothing about whether that regimen works, and it does not make the trial larger or better controlled than it was. Several things deserve emphasis, because they are systematically obscured in secondary content: The famous stroke trial is smaller and older than usually claimed. The Gusev/Skvortsova acute stroke study that generates the 12 mg and 18 mg figures enrolled 30 Semax-treated patients against 80 controls, and was published in 1997[11]. It is routinely described online as a ~200-patient multicentre trial from 2001. It was not. PubMed types it as a controlled clinical trial rather than a randomised controlled trial, the study is Russian-language, and no independent replication exists. These are acute stroke doses, in hospital, for days. The 12–18 mg/day figures come from patients with acute cerebral infarction receiving combined intensive therapy. They are not general-purpose doses, they were not studied in healthy people, and treating them as a template for anything else is an extrapolation the source does not support. The 2018 study’s 6,000 mcg/day is not the same thing as the 1997 study’s 12 mg/day. Different population (rehabilitation, not acute), different design, and half the daily amount. The abstract does not state which concentration was used, which means the volume administered cannot be derived from it. If it were the 1% solution, 6 mg/day would be 0.6 mL/day; if the 0.1% solution, 6 mL/day — an implausible nasal volume, given that the entire registered 0.1% bottle holds 3 mL. The 1% strength is the more likely reading, but this article will not assert it as fact, because the source does not. There is no established human dose for cognitive enhancement in healthy people. None of the above is a cognitive-enhancement protocol. The two healthy-volunteer studies that exist are neuroimaging studies with n=24 and n=52 measuring functional connectivity changes over 5–20 minutes[13][14] — they report that something measurable happens to brain networks, not that cognition improves, and neither was designed to establish a dose. Rodent work clusters at far lower per-kilogram amounts than the human clinical figures. Intranasal Semax at 50 and 250 µg/kg raised basal forebrain BDNF[9]; chronic intranasal work used 0.05 mg/kg (50 µg/kg) daily for 10–14 days[15]; the 2024 chronic-stress study used 60 nmol/kg intraperitoneally, which at 813.9 g/mol works out to roughly 49 µg/kg[5]. By contrast, 12 mg/day in a 70 kg adult is about 171 µg/kg/day. Interspecies dose translation is not a matter of matching mg/kg — allometric scaling, differing nasal anatomy and surface area, differing mucociliary clearance rates, and differing disease context all intervene — so this comparison is offered as orientation, not as a conversion. The two vial sizes most commonly encountered in the research-supply market are 5 mg and 10 mg; our reference pages break each down individually — the Semax 5 mg vial dosage protocol and the Semax 10 mg vial dosage protocol — and the tables below reproduce that arithmetic in full so it can be checked independently. Note throughout that this is unit-conversion arithmetic describing what a given dilution contains; it is not a recommendation that any quantity be administered, and the route with an actual evidentiary precedent is intranasal, not injection. A lyophilised vial contains a stated mass of peptide. Adding a volume of bacteriostatic water produces a concentration. That concentration, divided across a U-100 insulin syringe’s 100 units per millilitre, gives a mass per unit. The arithmetic is fully checkable: Concentration (mg/mL) = vial mass (mg) ÷ water added (mL) Per unit (mcg) = concentration (mg/mL) ÷ 100 units × 1,000 mcg/mg = concentration (mg/mL) × 10 Worked once, explicitly: a 5 mg vial with 2 mL of water gives 5 ÷ 2 = 2.5 mg/mL. One unit is 1/100 of a millilitre, so one unit holds 2.5 ÷ 100 = 0.025 mg = 25 mcg. The shortcut — multiply mg/mL by 10 to get mcg per unit — follows directly and is used for every row below. 1 mL 5 mg/mL 50 mcg 0.5% 1.5 mL 3.33 mg/mL ≈33.3 mcg ≈0.33% 2 mL 2.5 mg/mL 25 mcg 0.25% 2.5 mL 2 mg/mL 20 mcg 0.2% 5 mL 10 mcg 0.1% — matches the registered nasal strength[2] 100 mcg 1% — matches the higher registered strength[3] 4 mg/mL 40 mcg 0.4% 10 mL 0.1%[2] Notice what the two “0.1%” rows require: 5 mL of water into a 5 mg vial, or 10 mL into a 10 mg vial. Typical lyophilisation vials used for research peptides hold 2–3 mL of total volume. Neither of those additions physically fits. A great deal of Semax content blithely describes adding 5 mL to a 5 mg vial without noticing that the vial cannot hold it. The consequences are worth stating plainly. Reproducing the registered 0.1% strength from a small vial requires either transferring to a larger sterile container — which introduces a sterility and accuracy problem of its own — or accepting a higher concentration and adjusting the volume administered accordingly. Reproducing the 1% strength, by contrast, is trivial: 10 mg vial + 1 mL = 10 mg/mL = 1%, and 1 mL fits comfortably. This is a real constraint that follows from the arithmetic, and it is covered in more general form in our peptide reconstitution guide. The reconstitution dosage calculator will run these conversions for any vial size and water volume. 1 unit 2 units 4 units 10 units 200 mcg 8 units 20 units 300 mcg 3 units 6 units 12 units 30 units 400 mcg 16 units 40 units 600 mcg 24 units 60 units 750 mcg 7.5 units 15 units 75 units 1,000 mcg (1 mg) 100 units Note the measurement-error trade-off embedded in this table. At 100 mcg/unit, a 100 mcg target is one single unit — and a half-unit misread is a 50% error. At 10 mcg/unit the same target is 10 units, where a half-unit misread is a 5% error. More dilute preparations are more accurately measurable; they also consume vial volume faster and, per the capacity constraint above, may not fit. Our guide to reading insulin syringe units for peptides covers this trade-off in general. It is also worth stating that the “5 mg” on a research-chemical vial label is an unverified claim. Fill accuracy, actual peptide content versus gross mass including counter-ions and residual solvent, and purity are all assumptions. Peptide content by mass is frequently well below the labelled figure once trifluoroacetate counter-ions and water of hydration are accounted for, and that discrepancy is invisible without a certificate of analysis and, ideally, independent testing. Every number in these tables is arithmetically correct conditional on the label being accurate, and that condition is not established for unregulated material. Lyophilised peptide stored cold and dry is the stable form; peptide in aqueous solution is the unstable form. Standard laboratory handling for lyophilised research peptides is: refrigeration for reconstituted material, freezer storage for long-term lyophilised material, protection from light and from repeated freeze-thaw cycles, and bacteriostatic water rather than sterile water where a preservative is wanted for multi-draw use. Our guide to storing peptides before and after reconstitution covers the general principles, and none of it is Semax-specific. What is worth saying specifically about Semax is that no peer-reviewed stability study of reconstituted research-grade Semax under domestic storage conditions was identified for this review. The Semax stability literature that is accessible concerns proteolytic stability in biological media — how the peptide resists enzymatic degradation in serum or tissue — which is a different question from chemical stability in a vial of bacteriostatic water in a refrigerator over weeks. Protease resistance in vivo says nothing about hydrolysis, oxidation, or aggregation in storage; the two failure modes are unrelated. Two structural features of the molecule are worth noting as chemically-motivated inferences rather than documented findings. Semax carries a methionine at position 1, and methionine is among the more oxidation-prone residues, susceptible to conversion to the sulfoxide on exposure to dissolved oxygen, light, or trace metals — a reason for caution about long-held reconstituted solutions. It also carries a histidine at position 3, and histidine-adjacent sequences are associated with metal-catalysed oxidation. Neither observation has been demonstrated to matter for Semax specifically in any study located here, and both should be read as reasons to be conservative rather than as established degradation pathways. By comparison, the registered product is a manufactured solution with a defined formulation, a manufacturer-established shelf life, and quality-control release testing behind it. A reconstituted research vial has none of those. Any “use within X days” figure attached to reconstituted Semax is a convention borrowed from general peptide practice, not a Semax-specific finding, and it should not be presented as one. This is where the search intent and the evidence diverge most sharply, so it deserves an unhedged answer: the intranasal route is the only one with a genuine human dosing precedent, and the asymmetry is not close. The registered Russian product is a nasal solution in both strengths[2][3]. Every human study cited above that specifies a route specifies intranasal[11][13]. The canonical rodent mechanistic work — the BDNF induction that constitutes Semax’s central mechanistic claim — used intranasal administration[9], as did the chronic behavioural work[15]. The rationale is nose-to-brain delivery: a route that may bypass the blood-brain barrier for a peptide that would otherwise cross it poorly. Very little, and this needs to be said clearly rather than softened. There is no published human dosing study of subcutaneous Semax that this review could identify. The parenteral evidence is preclinical and mostly intraperitoneal — the 2024 chronic-stress study used daily IP injections at 60 nmol/kg[5] — and the 2020 healthy-volunteer connectivity study describes “injection” without the abstract specifying subcutaneous[14]. Intraperiton