Skip to content
Peptide Evidence Registry

Tissue repair

BPC-157

The internet's favourite healing peptide: a mountain of rodent data and almost no controlled human evidence.

L3Early human studiesResearch / unapproved supplyIn clinical development
How settled this rating is:
very-low confidence in the current evidence rating
Routes studied:
Not established · Subcutaneous injection, Nasal, Oral (marketed but never approved)
Also known as:
Body Protection Compound 157, PL 14736, pentadecapeptide BPC 157
Last reviewed:
2026-09-29
Jump to section

Evidence passport

BPC-157

A one-screen evidence identity for this compound. These fields describe research maturity, confidence and legal status. They are not a safety guarantee, efficacy percentage or recommendation.

Reviewed 2026-09-29
Evidence level
L3Early human studies
Confidence
Very low confidence in rating
Regulatory status
Research / unapproved supply
Development status
In clinical development

Evidence strength

Human efficacy evidence26 of 100 evidence strength

Higher = more and better-quality human evidence. It does not mean the effect is bigger or that the compound works for you.

Human safety evidence18 of 100 evidence strength

Higher = more mature, better-documented human safety data. It does not mean the compound is safer than one with a lower number.

The 0-100 values rate the strength and maturity of evidence, not how effective or safe the compound is.

Evidence base
Mostly animal and laboratory research
Furthest established stage
Phase II
Human research maturity
Preclinical for healing claims, with uncontrolled human observational documentation and a recruiting Phase 2 trial with no results
Source curation
Sources fully curated

Routes recorded

SC injection· unapprovedNasal· unapprovedOral· unapproved

Evidence Diff

What changed since the previous recorded review?

Grade unchanged

Previous recorded review

L3

Latest recorded review

L3

Latest material update

Retained L3 for retrospective human exposure and symptom documentation, while musculoskeletal healing remains at L2 preclinical evidence.

The September 2026 preprint is uncontrolled and cannot establish causality. A registered Phase 2 trial has no posted efficacy results. Confidence remains very low. Reviewed 29 September 2026.

This comparison uses only recorded change-log fields. Historical confidence is not inferred when it was not explicitly stored. A grade change still requires human editorial review.

Most of this evidence comes from animals and cells

The supporting research for BPC-157 is predominantly preclinical (Level 1–2). That is a legitimate starting point for science and a poor basis for personal decisions. Human results are not simply a scaled-up version of mouse results.

Overview

Human efficacy evidence strength26 of 100 evidence strength

Higher = more and better-quality human evidence. It does not mean the effect is bigger or that the compound works for you.

Preclinical (animal / lab) evidence strength78 of 100 evidence strength

Higher = more animal, tissue or cell research. Preclinical results are a starting point, not proof of a human benefit.

Human safety evidence strength18 of 100 evidence strength

Higher = more mature, better-documented human safety data. It does not mean the compound is safer than one with a lower number.

Evidence for the assessed benefit

L2Animal / lab studiesVery low confidence in rating

Musculoskeletal healing: animal and laboratory findings; no adequately controlled human efficacy results.

Human research context: L3 describes retrospective clinical-note surveillance of human exposure and reported symptoms, not demonstrated healing efficacy.

Research maturity, benefit evidence, safety and approval are different questions.

How to read these scores

Every number here rates the evidence, not the compound. They are editorial 0-100 strength ratings, not percentages, and not a rating out of five stars.

  • Human efficacy evidence strength — Higher = more and better-quality human evidence. It does not mean the effect is bigger or that the compound works for you.
  • Preclinical evidence strength — Higher = more animal, tissue or cell research. Preclinical results are a starting point, not proof of a human benefit.
  • Human safety evidence strength — Higher = more mature, better-documented human safety data. It does not mean the compound is safer than one with a lower number. A score of 80 does not mean “80% safe”.
  • Evidence level (L0-L5) — the strongest type of study behind the compound, from anecdote (L0) to large replicated human trials (L5). It is a description of the study evidence, not a quality, safety or recommendation rating.
  • Confidence — how likely we think it is that this rating changes as new research lands. Low confidence means the picture is unsettled, not that the compound is bad.

None of these ratings is medical advice, a safety guarantee, or a recommendation to use anything. How we grade evidence.

How far human research has progressed: Preclinical for healing claims, with uncontrolled human observational documentation and a recruiting Phase 2 trial with no results

Routes & formulations

Marketed / research routes — never approved

SC injection· unapprovedNasal· unapprovedOral· unapproved

Regulatory material describes products marketed as injections, nasal sprays and oral capsules. That is a description of the unregulated market, not evidence that any route works or is safe. No approved formulation exists for any route.

Route availability is not evidence of efficacy, and evidence from one route does not transfer to another. The Registry never provides dosing or sourcing information.

Peptide lifecycle

  1. Discovery

  2. Preclinical

  3. Phase I

  4. Phase II

    Furthest established stage

  5. Phase III

  6. Approved

In clinical development

BPC-HAMSTR (NCT07437547), a randomised, double-blind, placebo-controlled Phase 2 trial in acute Grade II hamstring strain, began on 2 February 2026 and is recruiting. Registration is not evidence of benefit; no results are posted.

Last documented development: 2 February 2026 — BPC-HAMSTR Phase 2 trial started; estimated primary completion 14 February 2027

What it is

A synthetic 15-amino-acid sequence derived from a protein found in gastric juice. It has been studied for decades in rodents across an unusually wide range of injury models.

What people claim

  • Heals tendons, ligaments and muscle tears
  • Repairs the gut lining and helps IBD
  • Reduces inflammation and speeds recovery from almost anything

What the evidence actually says

The preclinical literature is large, consistent and mostly produced by a small number of related research groups. There is still no adequately powered controlled human efficacy result supporting the tendon and injury-recovery claims that dominate online discussion, although a 120-participant randomised placebo-controlled Phase 2 hamstring-injury trial is now recruiting. A September 2026 preprint (not peer reviewed) described 1,039 confirmed BPC-157 users identified from a federated US clinical-note network. That descriptive human dataset moves the strongest study type to Level 3, but clinical certainty remains very low: it is uncontrolled, retrospective, unstandardised and heavily confounded, so it cannot establish efficacy or causality.

Related evidence

Useful next reads based on this compound — not recommendations.

Human evidence

  • September 2026 preprint (not peer reviewed) — retrospective clinical-note surveillance across a federated US network of 15.2 million patients with notes: 1,536 notes mentioned BPC-157 and 1,039 users were confirmed.
  • Newly documented users per quarter rose roughly 33-fold, from 4 in Q1 2020 to 134 in Q1 2026.
  • Among the 205 users with a documented source, 36% reported compounding pharmacies and 35% grey-market peptide vendors.
  • 525 of 1,039 users (50.5%) had another therapeutic agent documented — testosterone 17.0%, NSAIDs 14.7%, TB-500 12.9%, corticosteroids 11.7% — so co-treatment is the norm, not the exception.
  • A directional symptomatic response was ascertainable for only 354 of 1,039 users. Within that selected subgroup, 279 of 354 (79%) were documented as improved, 5% worsened and 16% had no change. The 79% figure is not an efficacy rate for all documented users.
  • No directional response was ascertainable for 685 of 1,039 users (65.9%).
  • Critically, this retrospective observational design had no control group, no standardised outcome measurement, no reliable standardisation of product identity or dosing, substantial co-treatment, and selection, reporting and confirmation bias. It cannot establish efficacy or causality, and most outcomes were patient-reported.
  • BPC-HAMSTR (NCT07437547) is a recruiting randomised, double-blind, placebo-controlled Phase 2 trial in acute Grade II hamstring strain, with approximately 120 participants planned. It started on 2 February 2026, has no results posted, and has an estimated primary completion date of 14 February 2027. Trial registration is not positive efficacy evidence.
  • No adequately powered published randomised controlled trials supporting musculoskeletal healing claims.
  • Limited early-phase clinical activity historically explored in inflammatory bowel disease; results have not established efficacy.

Preclinical evidence

  • Numerous rodent studies reporting accelerated tendon, ligament, muscle and gut healing.
  • Reported protective effects against NSAID-induced gastrointestinal injury in animals.
  • Concentration of output within a small research network is a replication concern, not a conspiracy claim — independent replication is simply thin.

Animal and cell findings are Level 1–2 evidence.

Mechanism and pathways

  • Proposed upregulation of VEGF signalling and angiogenesis at injury sites.
  • Interaction with nitric oxide pathways.
  • Proposed effects on growth-hormone receptor expression in tendon fibroblasts.
  • All of this is mechanistic (Level 1-2) and does not establish clinical benefit.

Safety and unknowns

  • There is still no systematic prospective human safety dataset. Absence of reported harm is not evidence of safety.
  • The September 2026 clinical-note preprint documented neuropsychiatric symptoms in 1.4%, injection-site or dermatologic reactions in 1.4%, and gastrointestinal events in 1.3%. These are not incidence rates: retrospective notes may under-report adverse events, product exposure was not standardised, and the data cannot establish safety.
  • Among 129 patients with a documented reason for discontinuation, adverse effects accounted for 24 cases (19%). This selected subgroup does not provide an overall discontinuation or adverse-event rate.
  • FDA's compounding review process identified safety concerns serious enough to warrant advisory committee consultation — a regulatory safety concern about compounding, not proof of harm in any individual, and not a finding that the compound works.
  • Product quality is a real and separate risk: unapproved supply chains have documented issues with purity, sterility and mislabelling, and the 2026 dataset found most documented sources were compounding pharmacies or grey-market vendors.
  • Because it is injected by non-clinicians in most real-world use, infection and technique risks are underappreciated.

Cancer relevance

This section covers questions raised by how the compound works in the body, plus anything regulators and trial teams keep an eye on. It is not a statement that this compound causes cancer — for most compounds here, no human data exist in either direction.

Theoretical / mechanistic concern only

The most common concern is that promoting angiogenesis could theoretically support tumour growth. This is a mechanistic hypothesis (Level 1), not a demonstrated human cancer risk — no human data show BPC-157 causes cancer. Equally, no human data show it is safe in people with cancer. Both statements are true at once, and that is exactly the point.

Regulatory status

Regulatory decisions apply to specific jurisdictions, products, formulations and indications. An approval elsewhere is not automatically an approval in your country.

Check the regulatory note and linked primary sources for jurisdiction-specific details.

Not an approved medicine in any major region, and it remains widely sold as a 'research chemical'. On 23 July 2026, FDA's Pharmacy Compounding Advisory Committee voted 8-6 with one abstention to recommend BPC-157-related bulk substances for the 503A Bulks List, despite FDA staff's recommendation against inclusion. The FDA evaluated ulcerative colitis, not musculoskeletal healing. The committee vote is advisory and does not itself establish final FDA listing or drug approval. Advisory committee recommendations are non-binding: this is a compounding-policy process, not FDA drug approval, validation of efficacy claims or blanket legalisation. Our curated record does not establish a final FDA determination after that meeting. Compounding-policy update: FDA's nominations document updated on 22 April 2026 records that several nominated peptide substances were removed from the interim 503A Category 2 list because the nominations were withdrawn, with Pharmacy Compounding Advisory Committee (PCAC) consultation planned. FDA then held a PCAC meeting on 23-24 July 2026 covering these substances. That is an advisory and compounding-policy process only: it is not FDA approval, not a finding that any of these compounds work, and not a blanket legalisation. Our curated record does not establish a final FDA determination after that meeting.

  • 2026-07-23 · United States

    FDA's Pharmacy Compounding Advisory Committee considered BPC-157-related substances for possible inclusion on the 503A Bulks List. Committee recommendations are non-binding: this compounding-policy process is not FDA drug approval, proof of efficacy or blanket legalisation. BPC-157 remains unapproved.

Key study types

  • L2Animal / lab studiesRodent tendon and gut injury models

    Preclinical animal studies — Rats and mice

    Consistent healing effects in animals; concentrated within a small research network.

  • L3Early human studiesSeptember 2026 federated clinical-note surveillance preprint

    Retrospective observational clinical-note study; preprint, not peer reviewed — 1,039 documented BPC-157 users in a 15.2-million-patient US network

    Large descriptive real-world dataset, but uncontrolled, unstandardised and confounded; it cannot establish efficacy, causality or safety incidence.

  • L3Early human studiesBPC-HAMSTR Phase 2 trial (NCT07437547)

    Randomised, double-blind, placebo-controlled; recruiting, no results posted — Planned approximately 120 participants with acute Grade II hamstring strain

    Started 2 February 2026 with estimated primary completion 14 February 2027; registration asks an efficacy question but does not answer it.

Evidence grade

L3Early human studiesVery low confidence in rating

Preliminary human outcome research, including small or observational studies. Descriptive exposure studies alone do not demonstrate therapeutic benefit.

What would change our rating?

  • A published, adequately powered randomised controlled trial in a defined injury population.
  • Peer-reviewed publication of the 2026 clinical-note surveillance dataset, which is currently a preprint.
  • Independent replication of core preclinical findings outside the originating research network.
  • Any published human pharmacokinetic and safety dataset.

References and sources

Source transparency

A provenance snapshot of this profile. Counts show what kinds of sources are cited, not how much scientific weight each source deserves.

Sources fully curated
Last editorial review
2026-09-29
Recently reviewed · 17 days ago
Linked curated sources
7
Sources still pending
0
External specialist reviews
0
None recorded for this profile
Regulatory / label: 4Trial registry: 1Preprint — not peer reviewed: 1Sponsor / official: 1

Review provenance

Registry editorial review: Colin Jackson-White, creator and editor, who is not a doctor or clinician. External specialist review is credited only when a named reviewer has completed a review of this exact profile.

Literature-discovery and AI tools may help locate or challenge candidate evidence, but they are not counted as evidence sources here. Registry judgements are tied to the underlying papers, trial records, regulator documents and official study records.

Source hierarchy: regulatory documents and trial registries first, peer-reviewed work next, sponsor material flagged as non-independent. We never use supplier or clinic marketing as evidence, and where a verified source has not yet been curated we say so rather than inventing one.

Evidence change history

These entries record changes to our assessment of the published evidence — not changes to the compound itself.

  1. Retained L3 for retrospective human exposure and symptom documentation, while musculoskeletal healing remains at L2 preclinical evidence.

    Change 1 of 3
    L3Early human studies
    The September 2026 preprint is uncontrolled and cannot establish causality. A registered Phase 2 trial has no posted efficacy results. Confidence remains very low. Reviewed 29 September 2026.

  2. A September 2026 preprint documented 1,039 BPC-157 users in retrospective clinical notes, while BPC-HAMSTR (NCT07437547), a planned 120-participant randomised, double-blind, placebo-controlled Phase 2 trial in acute Grade II hamstring strain, is recruiting with no results posted.

    Change 2 of 3
    L3Early human studiesEvidence rating changed
    The rating moves from Level 2 to Level 3 only because the strongest study type is now uncontrolled human observational evidence. The reported 79% is 279 of 354 users with an ascertainable directional response, not all 1,039 users; 685 had no ascertainable response. The preprint cannot establish efficacy or causality, and trial registration is not positive efficacy evidence. Clinical certainty remains very low and no healing claim is proven. Reviewed 18 September 2026.

  3. Baseline entry: registry position set to preclinical-only for musculoskeletal healing claims.

    Change 3 of 3
    L2Animal / lab studiesBaseline entry
    Reviewed the available literature landscape. Preclinical breadth is large; controlled human evidence for tendon and injury claims remains absent, so confidence stays very low.

Related claims we’ve checked — in full

Claim verdict

Tissue repair

Not supported as wordedL2Animal / lab studies

“BPC-157 heals tendons in humans.”

This verdict applies to the exact claim above. It is not a verdict on the compound overall, a safety guarantee, or a recommendation to use it.

Registry verdict

Unsupported as stated — the human trials simply don't exist.

Why this verdict

  • The tendon-healing story comes almost entirely from rodent models (Level 2).
  • No adequately powered published randomised controlled trial in humans supports it.
  • 'Unsupported' means we don't know it works — not that we know it fails. Those are different claims and the difference matters.

Claim evidence map

Sources are mapped to this exact claim. Appearance here does not mean every source supports it.

2 mapped sources

Claim verdict

Safety

Not supported as wordedL1Theory only

“Research-grade peptides are the same as pharmaceutical-grade medicines.”

This verdict applies to the exact claim above. It is not a verdict on the compound overall, a safety guarantee, or a recommendation to use it.

Registry verdict

Unsupported — 'research use only' is a legal category, not a quality guarantee.

Why this verdict

  • Unapproved supply chains have documented problems with purity, sterility, mislabelled content and dose accuracy.
  • A certificate of analysis supplied by the seller is not independent verification.
  • This risk is separate from — and often larger than — the pharmacology of the molecule itself.

Claim verdict

Tissue repair

MisleadingL3Early human studies

“A new human study found BPC-157 worked for 79% of patients.”

This verdict applies to the exact claim above. It is not a verdict on the compound overall, a safety guarantee, or a recommendation to use it.

Registry verdict

Misleading — 79% describes a selected subgroup with a recorded response, not all documented users.

Why this verdict

  • The figure is 279 of the 354 patients whose clinical notes contained an ascertainable directional response, not 79% of all 1,039 documented BPC-157 users; 685 had no ascertainable directional response.
  • The preprint was retrospective and observational, most outcomes were patient-reported, product identity and dosing were not standardised, and co-treatments were common — including TB-500 in 12.9% of users.
  • With no control group, the study cannot establish that BPC-157 caused improvement or that a marketed Wolverine blend works.