One of my patients injected a peptide for several months. He had been promised better recovery and better health. Nothing happened. Eventually he sent an unused vial from the same supply to a laboratory.
No detectable active peptide. Only what appeared to be filler material. He had spent months injecting pharmaceutical fairy dust.
Then there is the other kind of case. In Australia, an accredited laboratory tested a black-market vial labelled as containing ten milligrams of retatrutide. It reportedly contained retatrutide, and almost nineteen milligrams of it (ABC News, June 2026). Nearly double the labelled dose.
So the underground peptide market offers a fascinating range of possibilities: nothing, something, or twice as much something as you ordered.
And yet thousands of intelligent people, including doctors, athletes and the entire population of Silicon Valley, are injecting these compounds and insisting that they work. So the question worth asking is not whether peptides are nonsense. It is narrower than that: which peptide, for what, at what dose, on what evidence, and made by whom?
Peptides are already real medicine
Let us dispose of one misconception immediately. Peptides are short chains of amino acids, and the body uses them as biological messages controlling appetite, metabolism, hormones, inflammation and tissue repair.
Insulin is a peptide hormone. Semaglutide and tirzepatide are peptide medicines. So is tesamorelin. Saying you do not believe in peptides is rather like saying you do not believe in tablets.
But peptide describes a chemical family. It is not a certificate of effectiveness, and it is not a certificate of nonsense either.
Where the craze actually came from
Silicon Valley did not invent these molecules. BPC-157 came out of Croatian research into gastric protection. Semax and Selank came from Soviet and Russian peptide programmes. MOTS-c was discovered by academics studying mitochondrial signals. GHK-Cu came from copper-binding peptide research into wound healing. Tesamorelin and retatrutide came through conventional pharmaceutical development.
What Silicon Valley contributed was a mindset: if the body has a pathway, perhaps we can optimise it. Then Ozempic and Mounjaro proved that a peptide could transform appetite and metabolic health, and people quite reasonably asked what else was out there. Overseas manufacturers noticed that consumers were willing to inject a molecule long before mainstream medicine was willing to prescribe it.
That is how a set of unrelated research compounds became one enormous category called peptides. Scientifically the grouping makes very little sense. Commercially it was genius.
Can this many people be wrong?
When I read peptide forums, I did not find one story. I found three. Some say it changed their life. Some think it helped but were also doing four other things. Some injected it, waited, and felt nothing whatsoever.
That is not what a simple scam looks like. It is what you would expect if several things were true at once: some molecules have genuine effects, some people respond better than others, some conditions improve by themselves, most users are also doing physiotherapy or dieting, product quality varies wildly, and expectation shapes symptoms.
Anecdotes are not trials. But they are not worthless either. In medicine, repeated patient observations are often the thing that generates the question a trial later answers. The trouble with forums is the missing denominator: you see the person whose shoulder finally improved, not the hundreds who felt nothing and quietly moved on.
The recovery peptides: BPC-157 and TB-500
BPC-157 is the celebrity of this world, nicknamed the Wolverine peptide. In animal studies it appeared to influence tendon, ligament, muscle, nerve and gut repair. You can see the attraction. If you have had tendon pain for a year and physiotherapy has not fixed it, a small peptide with promising animal data is tempting.
The formal human research is thin. One frequently cited pilot involved twelve women with interstitial cystitis, with the peptide instilled into the bladder, uncontrolled and without a placebo (PubMed 39325560). It cannot tell us whether injected BPC-157 heals a human tendon.
The fair conclusion is uncomfortable for both camps. The evidence does not show that BPC-157 works for sports injuries. It also does not show that everyone reporting improvement experienced nothing pharmacological. It may accelerate repair, it may alter pain or inflammation without rebuilding tissue, or the improvement may belong to rehabilitation and time. That is not a verdict. It is a research question nobody has funded properly.
TB-500 is usually stacked with it. It is a seven-amino-acid fragment related to thymosin beta-4, a 43-amino-acid peptide. A fragment may keep the important biological effect, or it may behave quite differently. A chapter can contain the best sentence in a book and still not be the book.
Because the two are taken together, alongside rest, physiotherapy and better protein intake, nobody can say which component deserves the credit. The FDA review of TB-500 found no human clinical evidence that it improves wound healing. That does not settle the biology. It tells us the consumer experiment has run far ahead of the clinical one, which is the defining feature of this entire movement.
MOTS-c: the mitochondrial peptide
MOTS-c has the most elegant story. It is a mitochondrial-derived peptide identified in 2015, and in animal models it influenced insulin sensitivity and physical performance (Lee et al, Cell Metabolism). For the longevity world, a molecule linking mitochondria, exercise and ageing was irresistible.
Users rarely describe a stimulant rush. They describe being less exhausted during exercise, or less foggy. Others notice nothing. There is a small human study showing that exercise increases the body’s own circulating MOTS-c (PubMed 33473109), which tells us it participates in human exercise biology.
It does not tell us that injecting it reproduces exercise. Your body temperature also rises when you train. We do not inject fever.
Semax and Selank: the peptides with real human data
These two deserve more respect than the internet gives them, and less certainty than their fans claim.
Semax is a seven-amino-acid peptide related to a fragment of ACTH, modified to keep the neurological effects without the hormonal ones. It is registered in Russia as an intranasal medicine. Selank is another seven-amino-acid peptide, derived from tuftsin, developed as an anti-anxiety treatment.
In a 1997 Russian study, Semax was given to 30 patients during acute ischaemic stroke and compared with 80 receiving conventional treatment, with faster recovery of some neurological functions reported (PubMed 11517472). A 2018 study followed 110 patients in stroke rehabilitation and reported higher plasma BDNF and greater improvement in daily function (PubMed 29798983). Small human imaging studies show both peptides produce measurable changes in brain activity (PubMed 32342318).
For Selank, a 2008 randomised study of 62 patients with generalised anxiety disorder or neurasthenia compared it against medazepam, a benzodiazepine, and reported similar anti-anxiety effects with less sedation (PubMed 18454096).
That is considerably more than rodents and testimonials. It is also small, mostly from one research ecosystem, compared against an active drug rather than placebo, and not independently replicated internationally. A change on an fMRI scan proves biological activity. It does not prove you will think more clearly.
One correction worth making, since it circulates constantly: the FDA’s 2026 review covered Semax, not Selank, and it concluded the evidence was insufficient to establish effectiveness for the uses reviewed (FDA briefing document). Anyone telling you the FDA recently cleared both is mistaken.
GHK-Cu: the beauty peptide
GHK-Cu is a copper-binding tripeptide with a long history in topical skincare research, and a reasonable scientific basis for investigating it on skin. Users report better texture, less redness, sometimes thicker-looking hair.
Injectable GHK-Cu is a different proposition entirely. The route changes exposure, dosing and risk, and topical evidence does not transfer. I have seen patients injecting it from overseas sellers. One used it for months with no improvement; laboratory testing found no detectable active GHK-Cu. That does not invalidate the molecule. It invalidates the assumption that a label proves what is in a vial.
When a peptide becomes actual medicine
Tesamorelin shows what the finish line looks like. It stimulates the growth-hormone pathway and is FDA-approved to reduce excess abdominal fat in adults with HIV-associated lipodystrophy, with pooled Phase 3 data in more than eight hundred participants (PubMed 20554713). Online, people use it well outside that population and track their results with DEXA scans, IGF-1 levels, or the highly sophisticated instrument known as how my trousers fit.
Evidence in one population does not transfer automatically to healthy adults chasing fat loss. It affects the growth-hormone and IGF-1 axis, which needs judgement and monitoring. It is a real medicine, not a wellness vitamin.
Retatrutide changes the argument further. In a randomised Phase 2 trial of 338 adults with obesity or overweight, the highest dose produced around 24.2% weight loss over 48 weeks, against 2.1% on placebo (PubMed 37366315). It is investigational, not approved, and the black market began selling it while the trials were still running.
A published structure is not a finished pharmaceutical. Reproducing it requires the right sequence, the modified amino acids in the right places, the fatty-acid chain attached correctly, truncated material removed, degradation controlled, consistent potency and sterile manufacture. Which brings us back to that Australian vial: the problem is not only that the black market may sell you nothing. It may also sell you something very powerful in an unpredictable quantity. Fairy dust is bad. Fairy dust with potent receptor activity is worse.
Why some of these will never be properly studied
Peptide advocates make a fair point here. Trials cost hundreds of millions, and a company will hesitate to fund one for a sequence that has been public for decades and cannot be protected. Potentially useful treatments can stay under-investigated simply because they are commercially unattractive.
But commercial neglect explains why the evidence is missing. It does not tell us what the missing evidence would have shown. The trial could be positive, negative, or positive in one narrow group. The honest conclusion is that we should fund independent research, not that we should pretend the research already happened. Otherwise profitable molecules get Phase 3 trials and unprofitable ones get studied through Reddit comments.
What the FDA vote actually meant
In July 2026 the FDA’s Pharmacy Compounding Advisory Committee considered seven peptide families and recommended six for possible inclusion on the Section 503A Bulks List (FDA meeting materials). The vote was non-binding. It was not approval, it did not establish safety or effectiveness, and it did not legalise anything currently sold online.
Supporters argue that a regulated compounding route would at least give patients identity, sterility, traceability and supervision, which beats an anonymous overseas seller. FDA scientists emphasised limited evidence and unresolved safety questions. Both positions are reasonable. The disagreement is really about how medicine should respond when widespread use arrives before adequate evidence.
Could it all be placebo?
No. Tesamorelin changes visceral fat in controlled trials. Retatrutide produces substantial weight loss. These are not expectations.
For the experimental end of the spectrum, expectation probably contributes to some reported improvement, particularly for subjective outcomes such as pain, energy, anxiety and concentration. But placebo is not lying. Expectation genuinely alters symptom perception, stress responses and behaviour.
The realistic explanation is a mixture: genuine pharmacology, natural recovery, expectation, concurrent treatment, lifestyle change, selection bias, product inconsistency and marketing, in proportions that differ between molecules and between people.
My evidence spectrum
Rather than sorting peptides into good and bad, I place them on a spectrum. At one end, approved medicines such as tesamorelin: defined manufacturing, controlled trials, a specific indication. Next, investigational pharmaceuticals such as retatrutide: strong data, development still underway. Then Semax and Selank: genuine human studies and clinical use abroad, awaiting independent replication. Then BPC-157 and MOTS-c: compelling biology, encouraging preclinical work, a strong user signal, very little reliable clinical evidence. GHK-Cu sits in two places at once, plausible topically and sparse by injection. TB-500 sits closest to the speculative end.
Spectrums move. A peptide that is experimental today may be established tomorrow, and another may fail once someone finally tests it properly. That is how science works.
So do they work?
Some definitely do. Some plausibly might. Some will work only for narrow conditions or particular patients. And some will turn out to be expensive dreams.
I do not think people using experimental peptides are foolish. They are responding to real problems: chronic injuries, fatigue, anxiety, ageing, poor recovery, and treatments that feel slow or incomplete. Some are early adopters of tomorrow’s medicine. Some are improving because of rehabilitation and time. Some are experiencing placebo. Some are taking a genuinely active molecule that has not had the research it deserves. Most likely, all of it is happening at once.
What we need is better evidence, and knowing what is actually inside the vial. Because we do not know yet is not the end of a scientific conversation. It is the beginning.
And please remember: research use only is not a medical qualification.
Watch or listen to the full investigation
Prefer to listen? The audio version is on Apple Podcasts and Spotify.
If you are considering any of this, the honest starting point is an assessment rather than a vial. My team and I see patients at SW1 Clinic in Singapore, and you can book a consultation here. I have also written about what the evidence says about collagen supplements and the muscle problem nobody discusses with GLP-1 medications, which come from the same instinct: read the study before you buy the story.
This article is general education and does not replace individual medical assessment. Do not use an injectable or prescription product without supervision from an appropriately qualified doctor.
Get these in your inbox. I write a free newsletter on ageing well, on the treatments that earn their place and the ones that do not. Subscribe here.
