BPC-157: What the Human Evidence Actually Shows (2026)

BPC-157 is searched about 190,000 times a month in the United States. It is sold as a research chemical, discussed as a tendon and gut repair peptide, and bought by people with an injury that has not healed. This piece is not about whether it works. It is about what is actually known, which is a smaller and stranger thing than either its promoters or its critics tend to say.

What is BPC-157?

It is a synthetic peptide fifteen amino acids long, originally isolated from gastric juice. That description comes from a 2025 narrative review in Current Reviews in Musculoskeletal Medicine, which calls it "a synthetic pentadecapeptide originally isolated from gastric juice" — a useful source here because its authors are independent of the group that produced most of the underlying research.

Does BPC-157 work in animals?

Yes, and the honest version of this story starts there rather than skipping it. The animal work is thirty years deep and it is not vague. In a 2003 rat study the Achilles tendon was cut through, treated, and then pulled until it failed: the treated tendons held more load than the controls. A 2006 study detached tendon from bone and reported faster reattachment, and also found BPC 157 blunted the damage caused by a systemic corticosteroid. Another followed a fully transected quadriceps muscle for 72 days. An independent Taiwanese group later replicated a plausible mechanism, linking the effect to VEGFR2 activation and new blood vessel growth.

That body of work is why the compound has a following. Dismissing it as snake oil misreads the literature just as badly as selling it as proven.

How many human studies of BPC-157 are there?

Three. Not three trials — three pilot studies, and the independent 2025 review states it plainly: "Only three pilot studies have examined BPC-157 in humans, including its use for intraarticular knee pain, interstitial cystitis, and intravenous safety/pharmacokinetics."

Their sizes are 16, 12 and 2. The knee-pain paper is not even a trial in the ordinary sense — it is a retrospective review of 17 patient charts, 16 of whom were reached by telephone afterwards. The interstitial cystitis study had twelve participants and no control group. The intravenous safety study had two. Thirty analysed people is the entire human evidence base for a compound with 190,000 searches a month.

Who ran the human studies?

The same person, in the same journal. All three list Lee E as first author and all three appear in Alternative Therapies in Health and Medicine. The 2025 intravenous study states in its own methods that the infusions were performed at a private clinic in Florida. None of that makes the findings wrong. It does mean there is no independent human replication to check them against — the independent replication that exists, the Taiwanese VEGFR2 work, is in animals.

Are there any clinical trials of BPC-157?

Two have ever been registered, and neither has reported. NCT02637284, a phase 1 safety and pharmacokinetics study, was registered in October 2015; its status has read "unknown" ever since and no results were ever posted. The second is live: NCT07437547 is a quadruple-masked phase 2 trial in acute hamstring strain, sponsored by Hudson Biotech, with 120 people estimated. It started on 2 February 2026 and its primary completion date is 14 February 2027.

One caveat worth stating, because it is the sort of thing that gets used to catch people out: "no published randomised trial" is not the same as "no trial has ever happened". The Croatian pharmaceutical company Pliva ran trials of a BPC 157 formulation for inflammatory bowel disease in the 2000s, before trial registration was mandatory, and a randomised ulcerative colitis study from that era exists only as a conference abstract. Those trials were run. They were never published as papers, and their data are not available to read.

So should you take BPC-157?

This site does not tell anyone to inject an unapproved compound, and it does not publish a dosage page for one either — which is why you will not find one here, despite 11,000 people a month searching for exactly that. Nor is the opposite claim available: absence of trials is not evidence of harm, and two people in a pilot study is not evidence of safety. Both certainties are unearned.

What can be said is when the uncertainty ends. A properly masked, properly powered trial is running right now and reports in 2027. Until then the honest sentence is the boring one: promising in rats, effectively untested in people.

If you have a joint or tendon problem in the meantime, the useful question is what already has human trials behind it. That is a different and much better answered question, and it is the one our joint-health ranking exists to settle.