Does BPC-157 Work? The Evidence, Side Effects and Safety Data

Last verified: 20 August 2026

There is almost no human evidence that BPC-157 does anything. That is not a hedge or a call for more research. It is a description of the published record as it stands on 20 August 2026.

The entire human literature is one controlled trial that missed its endpoint by a route nobody buys the product for, a registered trial whose results were never reported, and three uncontrolled studies of two to seventeen participants, all from a single investigator. The animal literature is large and largely the work of one research group. In February 2026, STAT reported that of roughly 200 published BPC-157 studies, nearly all include Predrag Sikiric or Sven Seiwerth as an author, and that a Polish review team warned this "could lead to confirmation bias."

In July 2026, an FDA advisory committee nonetheless voted 8-6-1 to recommend BPC-157 for compounding, over the objection of FDA's own review staff. Both things are true, and this article covers both.

What this means for you

If you are deciding whether BPC-157 is worth taking, the honest framing is that you would be the experiment. There is no dose shown to work, no route tested for the uses it is sold for, and no safety database large enough to detect an uncommon problem.

Three questions get conflated constantly here. Does the evidence support it? No. Is it lawful to compound? Not today — see our BPC-157 legal status piece and our peptide legal status overview. Is it prohibited in sport? Yes, at all times. Those axes move independently, and an advisory vote changes none of them.

What BPC-157 actually is

BPC-157 is a synthetic pentadecapeptide: a chain of fifteen amino acids made in a laboratory. It is described as a partial sequence of a larger protein, "body protection compound," said to occur in human gastric juice.

That description carries much of the marketing, so it is worth slowing down on. The claim that the parent protein exists in gastric juice, and that this fifteen-residue stretch is its active portion, originates with the same Croatian research group that produced most of the literature. It describes provenance. It does not demonstrate that the peptide does in the body what the parent protein is supposed to do. And synthetic BPC-157 is not isolated from anyone's stomach: it is manufactured, in the grey market by suppliers the buyer usually cannot identify.

FDA's 2026 review records that there is no established receptor and no established mechanism of action (FDA briefing document, BPC-157). The mechanistic stories you will read — nitric oxide signaling, VEGF-driven blood vessel formation, growth factor modulation — are inferences from animal and cell experiments. They describe things that changed in rodents alongside the peptide. They do not identify what it binds to in a human being. The only human pharmacokinetic data FDA could find is a 2003 abstract in which BPC-157 was not detected in plasma. That used a rectal route and cannot be extrapolated to injection, but it is the only human absorption data there is, and it is a null result.

What it is marketed for

BPC-157 is marketed for tendon and ligament healing, gut repair, post-surgical recovery and general injury recovery. Some sellers extend that to joint repair, "leaky gut" and neuroprotection.

We are describing what sellers claim, not endorsing it. Under the FTC's substantiation standard an efficacy claim requires competent and reliable scientific evidence, and for BPC-157 that evidence does not exist. Note too that the only indication ever nominated to FDA was ulcerative colitis (FDA advisory committee agenda). Tendon healing, the most common thing it is sold for, was not the nominated use.

The human evidence, in full

This is the whole of it. FDA's 2026 briefing document is the most complete synthesis of human BPC-157 data in existence, and it fits in one table.

STAT describes the same record from another angle: an unpublished ulcerative colitis trial run by PLIVA, the Croatian company that developed the compound; a 2015 trial whose data were withdrawn before external review; and three uncontrolled studies of two to sixteen participants by a single Florida endocrinologist, published in Alternative Therapies in Health and Medicine.

The two accounts line up, with two discrepancies we could not resolve. FDA's table puts the largest uncontrolled study at 17 participants where STAT gives the range as two to sixteen. And we could not confirm whether FDA's "Ruenzi et al., 2005" is STAT's unpublished PLIVA trial, or whether the registered Phase 1 that never reported (NCT02637284) is the 2015 trial STAT describes. We flag the gaps rather than paper over them.

What "uncontrolled, n=2 to 16" means

If you do not read clinical literature for a living, that phrase does most of the work in this article and deserves plain explanation.

Uncontrolled means there was no comparison group. Everyone got the peptide. Nobody got a placebo. So when participants improved, there is no way to know what they would have done without it. Injuries heal on their own. Symptoms fluctuate and regress toward the mean. People who have just paid for a treatment and are seen regularly by a clinician report feeling better for reasons unrelated to the drug. An uncontrolled study cannot separate treatment from time, from placebo, or from the natural course of the condition. It records that things changed. It cannot say why.

No blinding means expectation is in the results. Participants knew what they were getting, and so did the investigator assessing them. In pain and recovery research, where outcomes are subjective and self-reported, that is exactly where apparent effects come from when they are not real.

Two to sixteen people is not a sample. It is a handful of individual cases. At that size, two good outcomes look like a pattern and two bad ones look like bad luck, and neither can be distinguished from noise.

None of this implies the investigator did anything improper. Case series legitimately generate hypotheses worth testing. They are simply not a way to answer whether a treatment works, and "there are human studies" implies something to most readers that these do not deliver.

Note the route column, too. The only controlled trial used an enema. FDA states the consequence plainly.

"We found no studies that administered BPC-157 to humans via the proposed oral, SC, nasal, or transdermal route of administration." — FDA briefing document, BPC-157, July 2026

The animal literature and the one-group problem

The animal literature is genuinely extensive: roughly 200 published studies, overwhelmingly in rats and mice, reporting benefit across an unusually wide range of models — tendon injury, gut lesions, wound healing, nerve damage.

The concentration is the issue. Per STAT's investigation, nearly all of those studies include Predrag Sikiric or Sven Seiwerth as an author, and a Polish review team examining the body of work warned it "could lead to confirmation bias."

Here is why single-group concentration matters even when everyone involved is scrupulous. Every lab has its own animal strains, injury models, assays, and habits about which experiments get written up and which get abandoned, and those choices are invisible in a published paper. When one group produces nearly all the data, there is no way to tell which reported effects are properties of the molecule and which are properties of that lab's methods. An independent group repeating the work with different animals, different reagents and no stake in the outcome is the only test that separates the two. For BPC-157, that test has largely not happened.

A quieter problem sits alongside it. A field dominated by one group tends to publish positive results, because negative findings from the lab that champions a compound rarely get written up. The literature can look uniformly favorable while the reality is mixed. That is what confirmation bias means here. And separately: rodents are not people, and animal evidence is where drug development starts, not where it finishes.

What FDA's reviewers concluded

FDA's review team assessed BPC-157 ahead of the July 2026 advisory committee meeting. Its conclusions are unambiguous.

"There is insufficient clinical safety information to characterize the safety profile of BPC-157." — FDA briefing document, BPC-157

Across the peptides under review, FDA's reviewers found the studies, as reported by STAT, to be "short in duration, small in sample size, and insufficient to establish safety or effectiveness." The team recommended against including any of the seven peptides on the agenda, in either free base or acetate form, on the 503A bulks list (FDA briefing document, introduction).

Outside the agency the assessment is blunter. Flynn McGuire, chief medical resident at the University of Utah, told STAT that BPC-157 "should not be used by humans." The US Anti-Doping Agency states that "it is unknown if there is a safe dose, or if there is any way to use this substance safely."

Safety: what is and is not known

Search results for "BPC-157 side effects" are dominated by pages reporting that it has few or none. That claim needs to be read correctly.

The named studies in FDA's table account for roughly 110 participants, a count that includes the placebo arm of the only controlled trial, with no long-term follow-up of any of them. By the standard statistical rule of thumb — the rule of three — if about 100 people were exposed and nothing was observed, a side effect affecting as many as one user in thirty could sit entirely undetected in a record that size.

What the toxicology does show

• Genotoxicity tests were negative. Ames, chromosome aberration and micronucleus assays support FDA's statement that "BPC-157 is not a mutagen." A real, if narrow, reassurance.
• Repeat-dose animal studies found changes. In 28-day intramuscular studies in rats and dogs, FDA records altered clotting — activated partial thromboplastin time shortened in rats, prolonged in dogs — plus increased ALT, glucose and triglycerides in both species. Liver and metabolic signals, in animals, of unknown human relevance.
• No carcinogenicity studies exist. None. The question has not been asked.
• Immunogenicity is flagged as a significant risk, amplified by peptide aggregation and impurities. FDA notes such reactions range from "antibody responses with no apparent clinical manifestations to life-threatening and catastrophic reactions," and are "often unpredictable."

The oncological question

The proposed mechanism for BPC-157 involves promoting the formation of new blood vessels. Solid tumors grow by recruiting new blood vessels. That overlap underlies a theoretical concern that BPC-157 could accelerate an existing tumor, raised in STAT's coverage and discussed at the advisory committee.

Two things should be said about it. It is a hypothesis derived from mechanism, not an observed harm: no case of BPC-157 promoting a cancer has been reported, and anyone telling you the peptide causes cancer is going beyond the evidence. But the reason nobody has observed it is that nobody has looked. There are no carcinogenicity studies, no long-term follow-up, and total documented human exposure of about a hundred people over short periods. A mechanism-based concern that has never been tested is not the same as one tested and dismissed. It sits open.

Product quality: what you are actually buying

Most real-world risk here is not pharmacology. It is manufacturing.

BPC-157 is not an approved drug and cannot lawfully be compounded today, so essentially all consumer supply comes from grey-market sellers labeling product "research use only" — a disclaimer FDA treats as legally worthless where marketing shows human-use intent (21 CFR 201.128; see FDA's 17 June 2026 warning letter to Wholesale Peptide). Nothing independent verifies what is in the vial.

USADA's Matthew Fedoruk put the consequence about as plainly as it can be put: "It could be a peptide. It could be a steroid. It could be something just like water."

We looked for a peer-reviewed independent purity survey of grey-market peptides and could not find one. Everything returned was vendor testing or affiliate marketing dressed as testing, which we do not treat as evidence. The closest rigorous work covers the channel rather than the molecule: Fabresse et al., Forensic Science International 2021, PMID 33838562 analyzed 110 products seized from the bodybuilding black market and found 33% substandard, 32% counterfeit and only 19% matching the original product. That study examined steroids and pharmaceuticals seized in France, not peptides sold as research chemicals in the US. It sets a base rate for the channel, and we offer it only on that basis.

BPC-157 also has a stability problem. FDA's review records that the lyophilized free base is stable only around three weeks at room temperature and requires storage below −18°C. Product shipped internationally without a cold chain is likely degraded on arrival, whatever was in it originally.

What a certificate of analysis does and does not tell you

Sellers offer a certificate of analysis as proof of quality. A COA has genuine regulatory meaning: FDA's interim compounding policy requires a valid one before enforcement discretion applies to any substance. But the document a grey-market vendor posts is not that, and what it omits matters more than what it contains.

This is not speculative. In its 2026 reviews of BPC-157, TB-500, MOTS-c, Semax and Epitalon, FDA found the certificates of analysis submitted in support of formal nominations lacked testing for peptide-related impurities, aggregates, bacterial endotoxins, bioburden, particle size and residual solvents. For BPC-157 free base, none was provided at all. If documentation submitted to a federal advisory committee looked like that, a PDF emailed by an anonymous seller warrants less confidence, not more.

"Because there is lack of information regarding potential impurities... and the potential of peptide aggregation, we cannot rule out the potential for immunogenicity associated with these impurities and peptide-related aggregates." — FDA briefing document

The tension worth naming

On 23 July 2026 the Pharmacy Compounding Advisory Committee voted 8-6-1 in favor of recommending BPC-157 for the 503A bulks list (McDermott Will & Emery; AJMC). It did so on the evidence base described above, and after FDA's own scientists recommended rejecting every substance on the agenda.

The reader deserves both facts together. A federal advisory committee looked at this record and a majority said yes. FDA's review staff looked at the same record and said no. The committee's roster drew scrutiny for including more practitioners who prescribe or produce peptides than previous rosters, per AJMC.

What the vote did not do is change anything. It is advisory and non-binding. The 503A bulks list at 21 CFR 216.23 still contains six substances and no peptides, unchanged since February 2019. Amending it requires notice-and-comment rulemaking, which has not begun. Buchanan Ingersoll & Rooney, on 18 August 2026, states that the votes "did not amend the 503A Bulks List" and that pharmacies "should not interpret the favorable votes as authorization to begin compounding." The last comparable cycle took over two years. And a vote about compounding eligibility was never a verdict on whether the compound works. Our BPC-157 legal status page walks through the gates that remain.

What would change our assessment

We are not committed to this conclusion. We are committed to the evidence, and the evidence could change. Here is what would do it.

• A randomized, double-blind, placebo-controlled trial, adequately powered, in humans. Not a case series, not an open-label cohort, not a chart review.
• By a route people actually use. Subcutaneous or oral, matching how the product is marketed. Another enema trial would not answer the question consumers are asking.
• A prespecified primary endpoint registered before enrollment on ClinicalTrials.gov, with results posted whether or not they are favorable. The registered Phase 1 that never reported is the pattern this would need to break.
• Objective outcomes. Imaging-confirmed tendon healing, endoscopic remission, measured function — not self-reported improvement from an unblinded participant.
• Independent investigators. Run and analyzed by a group with no authorship history on the existing literature and no commercial stake, then replicated by a second group. One positive trial from a new lab would be news. Two would be a finding.
• Real safety work alongside it. Carcinogenicity studies, longer-duration toxicology and immunogenicity testing on characterized material — the work that would let the oncological question be resolved in either direction.

Any one of those moves the needle. Together they would make BPC-157 a drug candidate rather than a marketing category. Until then the accurate statement is that the evidence does not exist, and we will keep reporting it that way.

Frequently asked questions

Does BPC-157 work?

Unknown, because the studies that would answer it have not been done. The only controlled human trial missed its endpoint, and every other human study is uncontrolled with two to seventeen participants. FDA's reviewers found the evidence insufficient to establish effectiveness.

Is BPC-157 safe?

Unknown. Roughly 110 people appear in the entire published human record, with no long-term follow-up. FDA's review states there is "insufficient clinical safety information to characterize the safety profile of BPC-157."

What are the side effects of BPC-157?

The human studies report headache and flatulence as the most common adverse events, from 24 volunteers using a rectal route. Repeat-dose animal studies showed altered clotting and raised ALT, glucose and triglycerides. Pages claiming "no side effects" describe a near-empty database, not a clean one.

Does BPC-157 cause cancer?

There is no evidence that it does and no reported case. There is a theoretical concern, because the proposed mechanism involves new blood vessel formation and tumors grow by recruiting new blood vessels. No carcinogenicity study has ever been run, so the question is open rather than answered.

Does BPC-157 heal tendons?

It is marketed for tendon and ligament healing. That rests on animal studies, mostly rodent and mostly from one research group. No human trial has tested it for tendon healing, which was not even the indication nominated to FDA. That was ulcerative colitis.

Did the FDA approve BPC-157 in 2026?

No. An advisory committee voted 8-6-1 in July 2026 to recommend it for the list of substances pharmacies may compound. That is non-binding, the rule has not been amended, and compounding eligibility is not drug approval in any case. See our BPC-157 legal status page.

Can I get BPC-157 from a compounding pharmacy?

Not lawfully as of 20 August 2026. BPC-157 has no USP monograph, is not a component of any approved drug and is not on the 503A bulks list, so it fails all three statutory conditions. Detail is in our peptide legal status overview.

Why are there so many BPC-157 studies if the evidence is weak?

Volume and independence are different things. Of roughly 200 published studies, nearly all include one of two authors, per STAT, and a Polish review team warned this could lead to confirmation bias. Most are animal work. Independent replication is what makes a finding durable, and it is what this literature lacks.

Is BPC-157 banned in sport?

Yes, at all times. It is named under WADA category S0 for non-approved substances, per USADA, and appears on the Department of Defense prohibited dietary supplement ingredients list (Operation Supplement Safety).

This is consumer research about how peptides are regulated and sold. It is not medical advice, and we are not your doctor. Nothing here should be used to start, stop or change any treatment. Talk to a licensed clinician who knows your history.