BPC-157 vs TB-500: what the evidence actually separates

Neither BPC-157 nor TB-500 has a completed trial showing that it does anything to a tendon, a ligament or a muscle in a person. For one of the two, no study reporting that it was given to a human being exists at all: FDA searched the literature and found no such article, and no case reports of human use either ¹. The comparison a reader wants, which of the two works better for an injury, has no measured answer on either side. What can be compared is how much has been looked at, in what species, and by whom.

Two molecules, one of them routinely misnamed

	BPC-157	TB-500
Length	15 amino acids	7 amino acids
Sequence	GEPPPGKPADDAGLV	LKKTETQ, acetylated at one end
Origin	partial sequence of a protein in human gastric juice	residues 17 to 23 of thymosin beta-4
Molecular weight	1419.5	889.01
CAS number	137525-51-0	885340-08-9

Those sequences and identifiers are as FDA recorded them when it reviewed both substances in 2026 ^{2 1}.

The identity problem belongs to TB-500 alone, and it is not a quibble about nomenclature. TB-500 is sold and described as synthetic thymosin beta-4, and FDA states that thymosin beta-4 and TB-500 are not the same substance ¹. The seven-residue acetylated peptide was identified in a marketed TB-500 product by mass spectrometry in 2012, so what sits in the vial is a measured fact rather than a labelling opinion ³. A reader who has been told that TB-500 is synthetic thymosin beta-4 has been told something a regulator contradicts in terms.

A second gap sits inside that one. FDA notes that wound healing is an effect normally associated with the sequence in its unacetylated form, and states that the unacetylated peptide's profile cannot be directly extrapolated to the acetylated form, because acetylation irreversibly alters a peptide's charge, hydrophobicity and size ¹. BPC-157 has no equivalent problem: the sequence studied is the sequence sold, whatever else is wrong with the product supply.

The human record: almost nothing against nothing

Five human studies of BPC-157 exist, and none of them measured healing in a tendon, a ligament or a muscle. Two were reported only as conference abstracts and never published as papers: a randomised placebo-controlled trial in 53 people with ulcerative colitis, whose estimated between-group difference on a disease activity index was 1.6 points with a 95 percent confidence interval from -4.84 to 1.62, an interval that includes zero, and a phase 1 tolerability study with 32 people randomised and 24 dosed ². The three that reached publication had no control group and 31 participants between them ^{4 5 6}. Four trials are registered and none has posted results ^{7 8 9 10}. What each one measured is set out in the review of BPC-157's human evidence.

For TB-500 the equivalent count is zero. One trial is registered, in stable atherosclerotic cardiovascular disease, recruiting, with adverse events as its primary outcomes ¹¹. It is not a recovery study and no results are posted.

So on human evidence the comparison is almost nothing against nothing. BPC-157 has more, and what it has does not show that it works.

The one experiment that compared them, in rats

The two have been tested side by side once. 32 male Sprague-Dawley rats, 12 weeks old and around 330 g, had the Achilles tendon transected and surgically repaired, then were randomly assigned to four groups of eight: control, BPC-157, TB-500, and both peptides together. The abstract does not say what the control group received, and the other three had four weeks of injections. Treatment was intraperitoneal for four weeks ¹².

Maximum load to failure, the force the repaired tendon withstood before breaking, was higher in both treated groups than in controls and reached statistical significance in the TB-500 group (p < 0.05). The study also scored the tissue under a microscope on two systems, Bonar and Movin, on which a lower score means less degenerative change and better collagen alignment. Total Bonar scores were significantly lower in the TB-500 group (p = 0.016). Total Movin scores were significantly lower in the TB-500 group (p = 0.017) and in the combined group (p = 0.040). The BPC-157 group's scores were numerically lower than control without reaching significance on either total. Collagen type I expression did not differ significantly between groups; collagen type III did ¹².

The group given both peptides is the only place in the published record where the combination has been examined, and the authors report that combined BPC-157 and TB-500 treatment did not confer additional benefits compared with either agent alone ¹².

On those endpoints, in those rats, TB-500 separated from control where BPC-157 did not. That is what the experiment found. The authors' own conclusion is broader than that: they report both peptides as associated with improved histopathological parameters and extracellular matrix organisation, with TB-500 additionally showing a significant biomechanical advantage at four weeks ¹².

Regulatory status: the same, except in sport

Neither compound is an approved drug and neither has a pharmacopoeial monograph. That matters because the 503A Bulk Drug Substances List is a residual route: a pharmacy compounding under that section may use a substance covered by a pharmacopoeial monograph, or one that is a component of an approved drug, and only failing both does the Bulks List apply. In 2026 FDA proposed adding neither substance to it, having evaluated BPC-157 for ulcerative colitis ² and TB-500 for wound healing ¹. Both were heard by the Pharmacy Compounding Advisory Committee on 23 July 2026 and no vote is published for either, so each stands as a proposal rather than a decision.

The one substantive regulatory difference is in anti-doping. BPC-157 is named in section S0 of WADA's Prohibited List, the class covering substances with no current approval by any governmental regulatory health authority for human therapeutic use, and is classed there as a Specified Substance. TB-500 is named in section S2.3, growth factors and growth factor modulators, where substances are non-Specified. Both are prohibited at all times rather than in competition only. The list was read on 2 October 2026 ¹³. Specified status does not mean a substance is less banned; it is a category the anti-doping rules use when weighing a sanction.

Both are sold under a common name that does not reliably identify one substance. FDA treats that as a safety risk in its own right, on the ground that inconsistent naming can leave a patient dosed with a different substance from the one a physician ordered ^{2 1}. An analysis of products bought online found the content of TB500 and TB1000 products not systematically consistent with their stated descriptions ¹⁴.

Safety, where the emptier column is not the safer one

	BPC-157	TB-500
Adverse event reports retrieved by FDA	three	zero
Food and supplement complaints	three	two, both about a blend containing BPC-157
Published case reports	none identified by FDA	none identified by FDA
Repeat-dose toxicology	28 days, in rats and dogs	none, in any species
Genotoxicity	Ames, chromosome aberration and mouse micronucleus, all negative	none
Carcinogenicity study	none identified	none
Reproductive toxicity	gestation days 6 to 15 in rats	none

The BPC-157 column is FDA's 2026 briefing document and its adverse event review ^{2 15}; the TB-500 column is FDA's TB-500 briefing document ¹. Having a toxicology package is not the same as passing it: in that 28-day work FDA records clotting time shortening in rats and lengthening in dogs, which it treats as clinically relevant signals pointing to opposite risks in people, alongside raised liver-associated markers in rats ². One of the three BPC-157 reports recurred when the person restarted the product. That product contained both BPC-157 and TB-500, so the reaction cannot be attributed to either one ¹⁵. The two columns also come from searches ending nine months apart, December 2025 for BPC-157 and March 2025 for TB-500.

What this comparison cannot do yet

One compound has a thin human record that does not demonstrate a benefit and a 28-day toxicology package in two species. The other has no reported human administration, no toxicology in any species, and is widely sold as a substance FDA says it is not. The one experiment that measured them against each other used 32 rats and called itself exploratory.

A comparison needs two measured quantities. On the endpoint people are actually asking about, healing in a person, there are none.