Healing Peptides: BPC-157, TB-500 and What Tissue Repair Research Actually Shows

Healing peptides in brief

Three numbers tell most of the story. How many people have taken BPC-157 in a study, how many have taken TB-500, and how many collagen trials sit behind tendons and joints.

Every figure on this page comes from a PubMed abstract or a regulator document listed at the end. No dose, schedule or protocol appears for any unapproved compound. Nothing here is a recommendation to use any peptide.

What this guide covers

How tissue heals, and where healing peptides are meant to act

Every injury heals in three overlapping stages. First comes inflammation, the first few days, when blood and immune cells flood in and clear the damage. Then comes rebuilding, roughly days to weeks, when new blood vessels grow and cells lay down fresh collagen. Last comes remodelling, weeks to months, when that new collagen is lined up and strengthened by load. Tendons and ligaments are slow at every stage because they have a poor blood supply.

Healing peptides are sold as a way to speed the second stage. BPC-157 is said to grow new blood vessels and help tendon cells survive and move into the gap. Thymosin beta 4, the protein TB-500 is cut from, holds actin, the cell's internal scaffolding, and helps cells migrate into a wound (PMID 22074294). In rats, both effects show up. Wounds re-covered 42 percent faster at day 4 on thymosin beta 4 (PMID 10469335). Cut Achilles tendons healed stronger on BPC-157 (PMID 14554208).

Collagen peptides work at the third stage from a different angle. They supply the raw building blocks of collagen. In one study, men took 15 g of vitamin C enriched gelatin an hour before exercise. A blood marker of new collagen doubled (PMID 27852613). Load tells the tendon where to put the collagen. The peptide just makes sure there is enough to build with.

Which healing peptide has been tested in which tissue

Here is the whole field in one grid. Each cell shows the best evidence for that peptide in that tissue. Read across the rows to compare peptides, down the columns to see where each one has been tested.

The pattern is hard to miss. BPC-157 is a column of grey with one dark red cell, a chart review of 16 knee patients (PMID 34324435). TB-500 is a column of black with one grey cell, the rat tendon study (PMID 42542926). Thymosin beta 4 has one red cell, for leg ulcers (PMID 20536470). Collagen peptides have four red cells, one each for tendon, muscle, knee and skin (PMID 30609761, PMID 39060741, PMID 29018060, PMID 37432180).

Peptides for healing, tissue by tissue

The grid gives the shape. This table gives the numbers behind each cell, so you can check any claim against it.

Two things stand out. Every injectable healing result in tendon, ligament and muscle is a rat or mouse result. And the human results that do exist for tendons, muscle and knees come from a cheap oral supplement. It was used alongside training, with modest effects and low certainty. That is not the story the vial labels tell.

Each healing peptide, in one paragraph

BPC-157

A 15 amino acid piece of a stomach protein. In rats it healed tendon, ligament, muscle and gut injuries across dozens of studies (PMID 14554208, PMID 20225319, PMID 20190676). In people, three uncontrolled studies with 30 patients, all poor quality (PMID 42885452). Placed in FDA Category 2 for compounding in 2023 and banned in tested sport. Full story on BPC-157 dosage explained.

TB-500

A 7 amino acid fragment of thymosin beta 4, first sold for racehorses and first studied by a doping lab (PMID 23084823). One healing study, in rats: 60 micrograms per kilogram a day for four weeks made repaired Achilles tendons stronger (PMID 42542926). No human study of the fragment. Lab tests of online TB500 products found they did not match their descriptions (PMID 36482504). Full story on TB-500 explained.

Thymosin beta 4

The full 43 amino acid protein. It has real human trials. It was safe at up to 1260 mg a day into a vein for 14 days in healthy volunteers (PMID 20536472). A phase 2 trial suggested faster leg ulcer healing (PMID 20536470), and eye drop trials gave mixed results (PMID 26056426). No trial in tendons, joints or muscle, and no approval anywhere.

Collagen peptides

Collagen broken into small fragments you swallow. The only healing peptide with randomised trials in tendons, muscle and knees. Across 19 trials and 768 adults, collagen with training improved tendon structure and strength a little more than training alone (PMID 39060741). In Achilles tendinopathy, collagen plus calf exercises improved scores faster than exercises alone (PMID 30609761). A food supplement, cheap, no side effects reported in the trials.

GHK-Cu

A three amino acid copper peptide. In lab and gene studies it raises collagen and elastin production and supports wound repair (PMID 29986520). Controlled human trials are thin and mostly cosmetic. Now also sold as an injectable, with no human injection trial behind it.

KPV

A three amino acid tail of a natural anti inflammatory hormone, alpha MSH. In skin cells in a dish it cut inflammation and cell death from fine dust (PMID 40073467). It is sold for gut and joint inflammation. No human trial has been published.

Peptides for joint pain, tendons and the questions people ask

Peptides for joint pain. The best human data is collagen hydrolysate. A 2018 meta analysis of 20 supplements for osteoarthritis found collagen hydrolysate among seven with large short term pain relief. No supplement held up at long term follow up (PMID 29018060). The only injectable with knee data is BPC-157. In a chart review of 16 patients, 14 reported relief by phone, with no controls, no scales and no dose reported (PMID 34324435).

Peptides for healing tendons. Collagen peptides plus a calf strengthening programme improved Achilles tendinopathy scores faster than the programme alone in a 20 patient crossover trial (PMID 30609761). Ten grams a day for 16 weeks made healthy Achilles tendons stiffer without changing their size (PMID 40623147). BPC-157 and TB-500 have rat tendon results only (PMID 14554208, PMID 42542926).

Can peptides heal a torn meniscus? No study has tested any peptide on a meniscus tear in people. The knee chart review lumped meniscus tears with arthritis and tendon problems and did not image anyone (PMID 34324435). A torn meniscus needs a diagnosis, not a vial.

What are the big 3 peptides? In healing forums it usually means BPC-157, TB-500 and GHK-Cu, sometimes sold as one blend. None of the three has a human trial in tendons or joints. The one controlled test of BPC-157 with TB-500, in rats, found the pair did not beat TB-500 alone (PMID 42542926).

Peptides for recovery after training. Collagen peptides with training improved 48 hour recovery of reactive strength after muscle damage, with low certainty (PMID 39060741). No injectable healing peptide has a human recovery trial.

Peptides for bone healing and after surgery. No published human study tests any of the injectable healing peptides on bone or post surgical recovery. A 2026 review of peptides in orthopaedics called the preclinical work promising and noted the lack of clinical trials (PMID 41490200).

Are healing peptides safe?

For the injectables, nobody knows, because nobody has run the study that would tell us. People quote a reassuring line: no adverse events in the BPC-157 human studies. That comes from 30 people followed up by phone with no control group (PMID 42885452). That is an absence of reports, not evidence of safety.

The theoretical worry comes from the mechanism itself. Both BPC-157 and thymosin beta 4 heal partly by growing new blood vessels. A compound that grows vessels in a healing tendon could, in principle, do the same where you do not want it. A 2026 review of unregulated injectable peptides lists pathological angiogenesis and carcinogenesis among the concerns for exactly these compounds (PMID 42757290). Nobody has shown it happens. Nobody has run the study that would rule it out. The general page on peptide side effects covers the sterility and injection problems shared by every mixed vial.

What the evidence says actually helps a tendon heal

If the goal is a tendon or joint that works again, the strongest evidence is not a peptide. It is load. Progressive strengthening is the treatment that tendon research keeps coming back to. The one positive collagen tendinopathy trial paired its supplement with exactly that: twice daily calf strengthening and a return to running programme (PMID 30609761). The difference between an irritated tendon and a degenerated one, and how to load each, is on tendonitis vs tendinosis.

Collagen sits on top of load, not instead of it. In the meta analysis, the gains came from collagen combined with training over weeks (PMID 39060741). In the gelatin study, the extra collagen synthesis came from taking it an hour before a short bout of exercise (PMID 27852613). The supplement supplies the material. The loading tells the tendon where to lay it down.

Five mistakes people make about healing peptides

1. Reading rat healing as human healing

BPC-157 and TB-500 have strong rat tendon results (PMID 14554208, PMID 42542926). Neither has a controlled human tendon study. Rats heal fast and are injected daily into the belly cavity.

2. Quoting thymosin beta 4 trials for TB-500

The leg ulcer and eye trials used the full 43 amino acid protein (PMID 20536470). TB-500 is a 7 amino acid fragment that has never been in a human study (PMID 23084823).

3. Ignoring the supplement with the data

Collagen peptides have randomised trials in tendons, muscle and knees (PMID 30609761, PMID 39060741, PMID 29018060). They are the healing peptide with human evidence, and the cheapest one.

4. Using a peptide instead of load

The positive collagen tendon trial paired the supplement with twice daily calf strengthening (PMID 30609761). No study supports a peptide replacing rehabilitation.

5. Trusting the label on the vial

Lab tests of online TB500 products found the contents did not consistently match their descriptions (PMID 36482504). The peptide you read about may not be the one you inject.

Verdict: strong in rats, empty in people, and the jar beats the vial

Judged by the papers rather than the ads, the injectable healing peptides are rat compounds. BPC-157 has dozens of positive animal studies and 30 humans in three uncontrolled reports (PMID 42885452). TB-500 has one rat tendon study and no human study (PMID 42542926). Thymosin beta 4 has human trials, but for leg ulcers and eyes, not tendons. The peptide with randomised human trials in tendons, muscle and knees is collagen, taken with training, with modest gains and low certainty (PMID 39060741).

For readers here, the practical conclusion is simple. Load the injury, give it time, and if you want a peptide, the one with human data comes in a jar. The deep pages are on BPC-157 and TB-500. The rest of the field is on types of peptides and growth hormone peptides. The syringe arithmetic for any vial is on the peptide calculator.