Follow the money

Why nobody's actually tested these.

The science behind most peptides is thin. That is not a verdict that they failed a test. For most of them, nobody could make money running the test. The evidence gap is an economics story, not a safety verdict. Untested is not the same as disproven, and it is not the same as safe.

The whole thesis in one comparison

Semaglutide (Ozempic)

$29.3B

in 2024, roughly 70% of Novo Nordisk's entire revenue. Patented, owned, exclusive into the 2030s. So it got studied to death across dozens of huge trials.

BPC-157

~30

humans, total, across every published study. The sequence is in the public domain. No patent worth pursuing, anyone can make it, so no company will pay to study it.

The molecules that got billion-dollar trials are the ones a company could own and sell. The ones with no human data are the ones nobody can. The evidence tracks the patent, not the biology.

The arithmetic of a trial

Running real human proof is not a science-fair project. It is a bet measured in tens of millions, with long odds, that takes a decade.

$11–53M

What a single Phase 3 trial costs, depending on the field.

6.7%

Of drugs that enter human testing ever reach approval. The rest are write-offs.

~10.5 yrs

Start to finish, and over 40% of Phase 3 programs never even file.

You only make that bet if, when it works, a patent lets you charge enough to recoup it. No patent, no recoupment, no trial. Every compounder would free-ride on data you paid for the second you proved it.

The crux: you can't own them

Most popular research peptides fail at the patent step. BPC-157 is a fragment of a protein in human gastric juice, studied in animals since the 1990s. MOTS-c is a peptide your own mitochondria make. TB-500 is a piece of a natural protein. Naturally occurring, long-published sequences are nearly impossible to claim as new compositions of matter, the strongest kind of patent. There are weaker fallbacks (a specific formulation, a method of use), but those do not stop a compounder from selling the bare molecule, and they do not support the pricing power that pays back a $100M+ program. A trial on an unpatentable molecule is, coldly, irrational for a for-profit sponsor. The trial that would generate the evidence is exactly the trial no rational company funds.

The exceptions prove the rule

The peptides that didget real trials and FDA approval are, every one, ownable branded products. Tesamorelin (Egrifta) is a patented analog approved on trials of 816 patients, with a company behind it. The GLP-1s got studied across dozens of massive outcome trials because semaglutide and tirzepatide are patented compositions of matter worth tens of billions a year. The lesson is not "GLP-1s good, peptides bad." It is that the presence or absence of evidence tracks the presence or absence of a patent.

So who profits from not knowing?

A whole ecosystem sells these right now, in their unstudied state. None of them has any reason to spend money proving they work, or that they don't.

Compounding pharmaciesMix and sell peptide prescriptions. No trial required, and a trial only adds cost and risk.

Telehealth platformsQuick online intake, script routed to a compounder. Volume is the business; evidence is not.

Longevity & wellness clinicsCash-pay injectable protocols. The trial that might say 'it does nothing' is pure downside.

Gray-market vendorsSell it labeled 'not for human consumption' to sidestep the rules entirely.

Influencers & affiliatesMonetize attention and referral codes. Certainty would kill the intrigue.

The parties who could fund a real trial (pharma) have no patent upside. The parties who profit today have every reason not to risk it. So nobody runs it.

The part that cuts both ways

This is not a cover-up. It's a blank.

Do not read any of this as "it works and big pharma is hiding it." There is no buried positive trial. The honest read of "no good human data" is we don't know, not "it's secretly proven." Spectacular results in rats are not human results, and the graveyard of drugs that cured mice and did nothing in people is enormous.

But the same economics that block efficacy trials also block dedicated safety trials and any organized adverse-event tracking. So "few reported harms" is weak reassurance, the same way "few reported benefits" is weak condemnation. Both columns of the ledger are mostly empty for the same reason: nobody funds the data.

The real, named concerns are not hypothetical: the FDA flagged immunogenicity, manufacturing impurities, and thin human safety data, and showed lab evidence that compounded lots can differ from commercial ones. Notice that several of those are manufacturing quality problems as much as molecule problems. A clean molecule and a dirty vial are not the same risk, and the current evidence cannot cleanly separate them.

Bottom line

For most of these, the honest status is: largely unstudied in humans, neither proven effective nor proven safe, with specific quality risks flagged and almost no surveillance to catch problems. That is genuinely uncomfortable. It is also accurate. It is not "dangerous," and it is not "safe."

See how each one grades →How we grade

Sources include the Tufts/DiMasi R&D cost work, the HHS ASPE clinical-trial cost report, Wong et al. on trial success rates, Novo Nordisk and Eli Lilly FY2024 filings, the MDPI BPC-157 patent review, and FDA Category-2 / PCAC materials. Regulatory status is moving fast in 2026; dated claims are re-verified against the live record.