Peptides for dopamine: the short answer
Search that phrase and you get lists. Eight compounds, ten compounds, ranked, each with a confident sentence about dopamine receptors or dopaminergic tone. They are largely written by people selling the vials, and they share one feature: none tells you what was measured. That omission is not incidental, because for most of these compounds nothing was.
The honest state of it as of September 2026: no peptide on any of those lists has a controlled human trial in which a dopamine endpoint was measured. Not a synthesis rate, not a release event, not a transporter occupancy, not a receptor density on imaging. What exists is rodent neurochemistry for two compounds, rodent behaviour for several more, human trials on unrelated endpoints for one, and inference for the rest.
That does not make the compounds uninteresting. PE-22-28 sits on an elegant piece of ion channel pharmacology, and kisspeptin has randomised human data most of this category can only envy. It does mean the question people are actually asking, whether any of this will make them feel motivated again, has not been tested. Take the brand name off the vial and the picture is a preclinical field with a marketing layer bolted on.
What raising dopamine means, and what a swim test measures
Dopamine is not a dial. At least five separable things could be done to that system. A drug could increase synthesis by pushing tyrosine hydroxylase, the rate-limiting enzyme. It could increase vesicular release, phasically in bursts or tonically as background level. It could slow clearance by blocking the dopamine transporter, which is what stimulants do. It could change receptor number or sensitivity at D1 and D2 family sites, usually over weeks and often downward. Or it could leave dopamine untouched and alter what happens downstream, through neurotrophic signalling and transcription involving BDNF and CREB. A compound that raises tonic dopamine and one that sharpens phasic bursts can feel like opposites.
Then there is the measurement problem, where most of these claims break. To say something changed dopamine you need microdialysis, fast-scan cyclic voltammetry, tissue assays of dopamine and its metabolites, or in a human, positron emission tomography with a displaceable radioligand. A forced swim test is none of those. It measures how long a mouse stops struggling in a beaker, and it was built to detect the acute signature of tricyclic antidepressants, which is why almost anything monoaminergic shortens immobility in it.
So when a paper reports reduced immobility and a product page turns that into a dopamine effect, a rodent antidepressant screen has been relabelled as a human neurotransmitter measurement. That substitution is the core problem here, and once you see it you can audit any of these lists in a minute.
PE-22-28: what the actual paper reports
PE-22-28 deserves its reputation, and the real story is better than the marketing. It is a heptapeptide, glycine-valine-serine-tryptophan-glycine-leucine-arginine, corresponding to residues 22 through 28 of spadin. Spadin is itself a fragment of the propeptide cleaved from sortilin, the sorting receptor encoded by SORT1, during processing in the Golgi. An endogenous fragment found, characterised and then shortened.
The work came from the Heurteaux and Borsotto group at the French national research centre in Nice, and the key paper is the 2017 Frontiers in Pharmacology report on shortened spadin analogues with better TREK-1 inhibition, in vivo stability and antidepressant activity. TREK-1 is a two-pore-domain potassium channel; blocking it makes neurons expressing it easier to depolarise. Deleting its gene produces mice that resist depression-like behaviour across five assays, the observation the programme was built to exploit.
The numbers are why people know the compound. PE-22-28 inhibits TREK-1 with a half-maximal inhibitory concentration near 0.12 nanomolar against roughly 40 to 60 nanomolar for spadin, a several-hundred-fold gain, and it is selective: no significant activity at TREK-2, TRAAK, TRESK or TASK-1, and no hERG effect at ten micromolar. Half-effect times for the shortened analogues ran from about 14 to 23 hours in mice against something on the order of six or seven for spadin, given intraperitoneally and by oral gavage.
The behavioural results were shortened immobility in the forced swim test and, after four days of sub-chronic treatment, reduced latency to eat in the novelty-suppressed feeding test. The same treatment increased labelled dividing cells in the hippocampus and raised PSD-95 in cortical neurons, indicating neurogenesis and synaptogenesis.
Where the PE-22-28 dopamine claim comes from
Nowhere in that literature is dopamine measured. The chain producing the claim runs like this: TREK-1 is heavily expressed in the dorsal raphe nucleus; blocking or deleting it increases firing there; therefore the peptide raises monoamines; therefore it raises dopamine. Every step is defensible until the last, where it collapses, because the dorsal raphe is the brain's principal serotonin nucleus. The demonstrated effect is serotonergic firing, not dopaminergic tone.
A quieter route to the same claim runs through the neurogenesis and PSD-95 data. Those are plasticity endpoints, in the same family as BDNF signalling, and it is not absurd that a compound changing spine density in cortex might eventually change how reward circuits behave. But that is a hypothesis two removes from dopamine, untested even in a rodent.
The regulatory position needs no hedging. PE-22-28 has no human trial of any kind, no registered study, no approval anywhere, and no pharmacokinetic or safety data in people. It circulates as research-use-only powder, meaning the seller has made no representation about sterility, identity or purity. A molecule with this selectivity and a clean rodent profile is exactly what ought to go into a phase 1 study, and nobody has.
Semax: the entry with real striatal monoamine data
Semax is where actual neurochemistry enters the list. It is methionine-glutamate-histidine-phenylalanine followed by proline-glycine-proline: the 4 to 7 fragment of adrenocorticotropic hormone with the hormonal activity stripped out, plus a C-terminal tripeptide that slows peptidase degradation. It has been a registered intranasal medicine in Russia since the 1990s for cerebrovascular, cognitive and optic nerve indications, and is unapproved in the United States, the European Union and the United Kingdom.
The dopamine claim here has a real citation behind it. A 2005 Neurochemical Research paper reported that semax activates dopaminergic and serotonergic systems in rodents. What was measured was tissue and extracellular monoamine metabolite content in rat striatum, the serotonin metabolite rising substantially over hours, with dopamine release assessed alongside amphetamine challenge. Read carefully, the serotonergic signal is the cleaner one, and the dopaminergic component looks more like modulation of evoked release than a rise in resting dopamine.
The human record is the familiar problem. Russian controlled studies exist in acute ischaemic stroke, including work reporting elevated plasma BDNF alongside faster recovery, and none has been replicated independently under modern randomised, blinded, placebo-controlled conditions. There is more clinical material here than for most nootropic peptides, and none of it would satisfy a regulator outside its home jurisdiction.
Selank: anxiolytic first, dopamine somewhere behind it
Selank comes from the same Moscow institute and is built the same way: the tuftsin sequence threonine-lysine-proline-arginine with a proline-glycine-proline tail for stability. It completed Russian phase 3 testing as a selective anxiolytic and is a registered prescription medicine there, unapproved elsewhere. Its central claim is anxiolysis without the sedation, tolerance and dependence of a benzodiazepine, tested only domestically.
The clinical evidence usually cited is a comparative study of roughly sixty patients with generalised anxiety disorder and neurasthenia, setting selank against medazepam on the Hamilton, Zung and clinical global impression scales. A real trial with an active comparator, and also small, unblinded by modern standards, single-country and unreplicated.
On dopamine, the rodent literature describes region-dependent changes in monoamine metabolism, including serotonin turnover in hypothalamus and catecholamine handling elsewhere, alongside indirect modulation of GABA-A receptor expression and inhibition of enkephalin degradation. The dopaminergic component is one thread in a diffuse pharmacology, measured in rat brain regions, far from anything a person would notice as motivation.
Dihexa: a retraction sits underneath the mechanism
Dihexa is the angiotensin IV-derived compound developed at Washington State University and marketed on rodent data suggesting it is orders of magnitude more potent than BDNF at forming synapses. Its proposed mechanism is potentiation of hepatocyte growth factor signalling at the c-Met receptor rather than direct agonism at anything.
That mechanism took a serious blow in April 2025, when the Journal of Pharmacology and Experimental Therapeutics retracted two papers from the programme: the 2012 report on developing angiotensin IV analogues as hepatocyte growth factor and Met modifiers, and the 2014 report claiming the procognitive and synaptogenic effects of those peptides depend on that system. The notices state that a university investigation found falsified or fabricated data and named those responsible. Those papers are the load-bearing citations for the mechanism the compound is sold on.
Dihexa has no dopamine data whatsoever, in any species, and no human data of any kind. It carries a risk its own mechanism creates, because c-Met is a well-characterised oncogenic driver and amplifying growth factor signalling there systemically is not neutral in a body that may be carrying occult neoplasia. Of everything here, dihexa has the widest gap between marketing confidence and evidence.
Cerebrolysin and noopept: volume of use, thin evidence
Cerebrolysin is not really a peptide in the sense the rest of this list is. It is an enzymatic hydrolysate of porcine brain protein, a mixture of low-molecular-weight peptides and free amino acids whose active constituents have never been fully defined. It is approved and widely used across parts of Eastern Europe and Asia and is not lawfully marketed in the United States. Unusually here, enough randomised trials exist that Cochrane has reviewed it twice.
The reviews are the point. In acute ischaemic stroke the Cochrane assessment found no benefit on death or dependency and flagged a signal of increased serious adverse events. In vascular dementia the reviewers allowed possible effects on cognitive and global function in mild to moderate disease but declined to recommend routine use, citing few trials, heterogeneous durations and short follow-up. That is what it looks like when a widely used product meets systematic review.
Noopept, or omberacetam, is a dipeptide-derived prodrug rapidly metabolised to cycloprolylglycine, an endogenous neuropeptide. It is a prescription medicine in Russia for cognitive disorders of vascular and post-traumatic origin, with a comparative trial against piracetam behind it, and its reported mechanisms are increased hippocampal BDNF and NGF expression plus modulation of AMPA and NMDA receptor function. Dopamine is not part of any mechanism claimed for it in the primary literature. In the United States it is neither an approved drug nor a lawful supplement ingredient.
Kisspeptin-10: the closest thing to a human finding
Kisspeptin is the outlier, and it shows what this category could look like if anyone ran the studies. It binds KISS1R, also called GPR54, on hypothalamic GnRH neurons and sits at the top of the reproductive hormone axis. It also acts directly on limbic structures in a way at least partly separable from the hormone cascade, and that property produced the human data.
The 2017 Journal of Clinical Investigation report was a randomised, double-blind, two-way crossover, placebo-controlled study in twenty-nine healthy young men, using functional MRI while participants viewed sexual, couple-bonding, negative and neutral images. Kisspeptin amplified limbic activation to sexual and bonding cues, enhanced responses in amygdala, caudate, putamen and thalamus during an olfactory task, and reduced negative mood. Two later randomised trials in JAMA Network Open extended this to thirty-two premenopausal women and thirty-two men with hypoactive sexual desire disorder.
This is real work: randomised, placebo-controlled, crossover, in humans, published in serious journals. It is also not a dopamine measurement. The endpoint is the blood-oxygen-level-dependent signal, a proxy for neural activity in structures that happen to be dopamine-rich. Caudate and putamen lighting up is consistent with dopaminergic involvement and does not demonstrate it, which would need a displacement study with a dopamine radioligand nobody has run.
The ones on the list that are not peptides at all
Two compounds turn up constantly on peptides for dopamine lists without being peptides. Bromantane is the first, and it has the most direct dopaminergic mechanism of anything here, which is why the category error matters. It is an adamantane derivative, a cousin of amantadine and memantine, developed in the 1980s at a Moscow pharmacology institute and used in Russia for neurasthenia. Its reported action is upstream, on tyrosine hydroxylase and dopamine synthesis capacity, rather than on release or reuptake. It also has a doping history: athletes tested positive at the 1996 Atlanta Olympics, the disqualifications were overturned because the substance was not named on the list, and it was added the following year.
L-tyrosine is the second, and that error is more basic. Tyrosine is a single amino acid, and a peptide requires at least one bond joining two residues, so one residue alone is definitionally not a peptide. It is the dietary precursor for dopamine synthesis, and the controlled human literature on it is mostly about performance under acute stressors, cold, noise, sleep deprivation, demanding multitasking, where depletion of catecholamine substrate is plausible. That is narrow and conditional, not evidence that supplementing tyrosine raises dopamine in a rested, well-fed person.
Naming both plainly is worth doing, because how these lists are built tells you how they were made. A compound gets included on the strength of an association with dopamine in search results, not on the strength of belonging to the class the list claims to cover.
How to read a dopamine claim
Four questions settle almost every claim in this category, and they can be asked of any product page in under a minute. First, what was measured? If the answer is a behaviour, a mood scale or an imaging signal, dopamine was not measured. Second, in what species and by what route? Intraperitoneal injection in a mouse and intranasal administration in a rat are not statements about an adult human.
Third, against what control, and for how long? A crossover design with placebo is worth an order of magnitude more than an open-label series, and four days of dosing in a mouse says nothing about what receptor populations do over months, which is where the unwelcome adaptations appear. Fourth, and most diagnostically, was a mood or behaviour result quietly relabelled as a neurotransmitter result somewhere between the paper and the page?
Applied to the seven compounds here, the audit resolves quickly. PE-22-28 has a well-characterised channel target, a clean rodent package and no human anything. Semax and selank have genuine rodent monoamine assays and clinical records that stop at one border. Dihexa has a mechanism whose foundational papers were retracted. Cerebrolysin has been systematically reviewed and did not fare well. Noopept never claimed dopamine in the first place. Kisspeptin has the best human evidence and it is about sexual brain processing. Anyone searching for the best peptides for dopamine is asking a question the evidence has not been assembled to answer.