TB-500 (thymosin beta-4 fragment): what it is, how it works, and what the evidence says

TB-500 is a seven-amino-acid peptide with the sequence Ac-LKKTETQ ^[1]. That sequence is borrowed, almost literally, from positions 17 through 23 of thymosin beta-4 (Tβ4), a 43-amino-acid protein ^[2] found in nearly every nucleated mammalian cell. Tβ4 is one of the most abundant intracellular proteins in the body. It's not exotic. It's not foreign. It's part of normal cell biology.

The confusion starts here. "TB-500" and "thymosin beta-4" are used interchangeably in marketing copy, on forum threads, by vendors. They're not the same thing. Tβ4 is the full-length endogenous protein. TB-500 is a fragment, synthesized to reproduce the actin-binding motif (LKKTET) that drives much of Tβ4's downstream activity ^[3]. The fragment is mechanistically active in vitro at low concentrations. Whether it behaves like the full protein in a living human body is a question almost nobody has answered with a registered clinical trial.

Worth flagging upfront: when you read a study about "thymosin beta-4 accelerating wound healing," that study almost certainly used full-length Tβ4, not TB-500. The extrapolation from one to the other is plausible. It's not proven.

The core biology is genuinely interesting. Tβ4 binds G-actin monomers and keeps them in a polymerization-ready pool, just below the threshold where they'd start spontaneously forming filaments. When a cell needs to move, toward a wound or a damaged vessel, that buffered pool releases, actin polymerizes fast, and the cell migrates without having to synthesize new actin from scratch. The LKKTET motif that TB-500 reproduces is the part of Tβ4 doing this work. Site-directed mutagenesis confirmed it: knock out the motif, lose the activity.

Angiogenesis is the second pillar. In a 2003 study using human umbilical vein endothelial cells and a chick aortic arch sprouting assay, the seven-amino-acid actin-binding motif produced angiogenic activity roughly equivalent to the full Tβ4 molecule at low nanomolar concentrations. Strip out any portion of the motif and the activity disappears. So at the level of endothelial cells in a dish, the fragment really does behave like the parent protein. That's the strongest mechanistic case for TB-500 as a discrete entity.

There's more. Tβ4 stimulates keratinocyte migration at concentrations as low as 10 picograms in Boyden chamber assays ^[4]. It suppresses NF-κB, which damps down downstream inflammatory cytokines like IL-1β and TNF-α ^[5], and that anti-inflammatory pathway operates independently of the actin-binding one. It also promotes collagen deposition and extracellular matrix remodeling. Stack all of this together and you get a coherent story: a peptide that helps cells migrate, helps vessels form, and dampens inflammation along the way.

The biology is real. The leap from biology to therapeutic effect in a human is where things get thin.

The animal data, especially in rodent wound-healing models, are the strongest part of the file. In a 1999 rat full-thickness wound study, topical or intraperitoneal Tβ4 increased re-epithelialization by 42% at day four and up to 61% at day seven compared to saline ^[6]. Wound contraction improved measurably. Collagen deposition went up. Angiogenesis went up.

That's a meaningful effect size in a controlled model. Not a rounding error.

Cardiac repair studies in rodents show Tβ4 can reduce scar volume after experimental myocardial infarction. Corneal healing models, mostly in rabbits, show accelerated epithelial closure. Hair follicle work in mice shows enhanced regrowth. The preclinical breadth is genuinely substantial, and it's why pharmaceutical interest in Tβ4 existed in the first place.

Safety in animals also looks clean. The rodent no-observed-adverse-effect level was established at a high dose threshold, several-fold higher than the therapeutic doses used in efficacy studies.

But two caveats apply, and they matter. First: virtually all of this work used full-length Tβ4, not the TB-500 fragment. Second: rodent wound healing translates to humans about as reliably as rodent cancer drugs do, which is to say, badly more often than not. The translation gap from "accelerates re-epithelialization in a rat" to "heals a torn tendon in a thirty-five-year-old recreational lifter" is enormous, and nobody has bridged it.

Here is the honest answer to the most important question.

TB-500, the fragment, has seen remarkably little formal human evaluation. Every human safety claim you'll see attributed to "TB-500" is borrowed, by extrapolation, from trials that used full-length thymosin beta-4 instead. Any standalone clinical investigation of the fragment itself remains, at best, in its earliest stages, nowhere near the kind of robust Phase II or Phase III data that would justify the confidence with which it's being marketed.

The Tβ4 human data, for what it's worth, is reasonably clean. A Phase I study in 54 healthy volunteers tested intravenous doses from 0.05 to 25 micrograms per kilogram, single and multiple-dose, and reported no dose-limiting toxicities and no serious adverse events ^[7]. A Phase II trial in acute myocardial infarction patients showed cardiac protection signals ^[8]. A Phase II in stasis and pressure ulcer patients showed wounds that healed closed meaningfully faster than standard care. A pediatric cardiac surgery study examined safety in infants. A Phase II ophthalmic program tested it for dry eye with acceptable tolerability ^[9].

That's the corpus. A handful of small-to-moderate Phase II trials, all with the full-length molecule, mostly never advanced to Phase III, none leading to approval.

So when someone says "TB-500 has been clinically tested and is safe," what they mean, whether they know it or not, is that a different but related molecule, given intravenously under medical supervision at carefully escalated doses, was tolerated by a few hundred patients across a handful of indications. That's not nothing. It's also not what most people self-injecting TB-500 from a research-chemical vendor are actually doing.

The peptide-community use cases cluster around a few themes, and the evidence base behind each one ranges from thin to nonexistent.

Tendon and ligament injuries. The most common use case. Torn rotator cuff, achilles tendinopathy, chronic elbow issues. The proposed rationale: fibroblast migration, collagen deposition, and angiogenesis at the repair site. The evidence: animal studies in different tissues, mechanistic plausibility, anecdote. No human tendon trial exists.

Muscle strains and recovery. The pitch is faster recovery from training-induced microtrauma. Evidence: zero direct human trials in athletes. Mechanistic story: plausible. What forum users report: variable, often confounded by simultaneous rest, NSAIDs, physical therapy, and BPC-157 stacking.

Joint pain and cartilage. Largely extrapolated from rodent work. No human cartilage data. This is the use case where the marketing runs furthest ahead of what's been measured.

Post-surgical healing. Mechanistically the most defensible. Wound healing is where the strongest preclinical data live. But the studies that exist used Tβ4, not TB-500, and were limited to specific wound types under medical supervision.

General "longevity" and "regenerative" use. Use cases exist; the data don't. This is where the category most resembles a research chemical wearing a wellness costume.

The pattern across all of these: a real mechanism, a real animal signal, no human evidence in the actual indication being treated, and a community that's collapsed the distance between the three.

The science supports almost nothing about how to dose TB-500 in humans. That has not stopped a remarkably consistent protocol from circulating on forums and vendor sites: a loading phase of roughly 4 to 10 mg per week, split across two to three subcutaneous or intramuscular injections, run for four to six weeks, followed by maintenance dosing of 2 to 5 mg every couple of weeks ^[10].

This protocol has no peer-reviewed clinical validation. None. It's not derived from a Phase I dose-finding study. It's derived, as best anyone can trace, from bodybuilding-forum convention that solidified sometime in the mid-2010s and got copied across vendor product pages.

For reference: the Phase I trials of full-length Tβ4 used intravenous doses in the microgram-per-kilogram range. The Phase II cardiac trial used high-dose IV weekly infusions over several weeks under hospital supervision. The forum protocols use subcutaneous milligram doses based on neither.

What users report, in aggregate: subjective improvements in healing and pain, often within two to three weeks. What's missing from that report: blinding, controls, placebo arms, objective imaging, and any honest accounting for regression to the mean in injuries that were going to improve anyway.

Short answer: the dosing protocols circulating online are improvisation, not science.

TB-500 is a peptide, which means digestive enzymes shred it on contact. Oral bioavailability is effectively zero. Capsules sold as "oral TB-500" are, at best, a waste of money and at worst a contaminated waste of money.

Subcutaneous and intramuscular injection are the routes used in research and the routes used by self-administrators. Both are mechanistically reasonable; both require sterile technique that most home users underestimate. Intranasal formulations exist commercially and are sometimes pitched as "needle-free," but absorption data for the TB-500 fragment via intranasal delivery in humans don't exist in the peer-reviewed literature. The molecule is also smaller than full-length Tβ4, which affects stability and likely shortens the functional half-life, meaning whatever's absorbed nasally may not stay around long enough to matter.

In practice: if TB-500 does anything in humans, it does it injected. Everything else is marketing.

The known risks from human clinical data, again all from full-length Tβ4 trials, are mild. Injection site reactions were reported in a small minority of subjects. Transient headache occurred in a modest proportion of participants. No serious adverse events attributed to the compound in Phase I or Phase II. No immunogenicity. No cardiovascular signals. No abnormal labs.

That's a clean profile. It also doesn't apply directly to TB-500, and it doesn't account for what changes when an unregulated injectable is administered by a layperson outside medical supervision.

Adverse event reporting systems capture voluntary, unverified accounts from real-world users of substances like TB-500, but the picture they paint is inherently messy, confounded by concurrent substances, contamination, and self-administration error. Such reports don't establish causation. They do suggest that real-world use looks considerably messier than trial data.

The theoretical risks are where the careful reader should focus. A pro-angiogenic, pro-migratory peptide that promotes vessel formation and cell migration is exactly the kind of compound you'd want to keep away from anyone with an undiagnosed cancer. Tumors grow by recruiting blood supply. Nobody has tested what TB-500 does in someone with a pre-existing malignancy. Nobody has tested chronic toxicity over years. Nobody has tested reproductive effects.

And then there's the manufacturing question. Research-chemical vendors operate outside cGMP. Independent third-party testing of TB-500 products has, in past industry audits of similar peptides, turned up wide variation in actual peptide content, bacterial contamination, and endotoxin levels above safe injectable limits. The product in the vial is not always what the label says.

Worth flagging: the biggest real-world safety risk with TB-500 is probably not the molecule itself. It's what's in the vial alongside it.

TB-500 is not FDA-approved for any indication, and there's no IND on file for human therapeutic use. The compounding picture has moved twice in 2026 and is worth getting right, because it is the part most often misreported.

In September 2023 FDA placed TB-500 in Category 2 of its interim policy on bulk substances for 503A compounding — the bin for substances the agency says present significant safety risks, which in practice meant licensed pharmacies could not prepare it. On April 15, 2026 the agency removed it, along with eleven other peptides, and referred the question to its advisory committee. That was not authorisation: it lifted a stated prohibition without putting the substance on the 503A positive list.

On July 23, 2026 the Pharmacy Compounding Advisory Committee voted to recommend TB-500 for that list, 8 in favour, 6 against, 1 abstention, with the free base and acetate voted separately and landing the same way. FDA's own reviewers had proposed rejecting it, citing a lack of evidence to evaluate effectiveness. The committee's advice is non-binding; FDA has issued no final determination, and adding a substance to the list requires rulemaking — a proposed rule, a comment period, a final rule — none of which has begun. Our report on the vote has the record.

The practical position today: not prohibited, not authorised. No licensed pharmacy can compound TB-500 under 503A, because it is not on the list and has neither a USP monograph nor status as a component of an approved drug.

It is, however, legally sold in the United States as a research chemical for in vitro and animal use. The "research use only" label is the legal mechanism that lets the supply chain exist. It's also the disclaimer that disappears the moment the product reaches a customer who intends to inject it.

WADA prohibits TB-500 at all times under Section S2 (Peptide Hormones, Growth Factors, Related Substances and Mimetics). The 2022, 2024, and 2026 Prohibited Lists all explicitly name "Thymosin-β4 and its derivatives e.g. TB-500." ^[11] Detection methods for the peptide in urine and plasma have been published ^[12]. Athletes subject to WADA-compliant testing, including Olympic-pathway, NCAA, and most professional leagues with formal anti-doping programs, face sanctions for any use, regardless of intent.

The honest summary: TB-500 is legal to buy and possess in most jurisdictions as a research chemical, illegal to market for human use, prohibited in sport, and — for now, notwithstanding a favourable committee vote — still without a usable compounding pathway. That's a narrower gray zone than the community sometimes acknowledges, and the July vote has not widened it yet.

These are the two peptides most often paired in recovery protocols, and the rationale for stacking is mostly mechanistic intuition.

BPC-157 is a 15-amino-acid sequence derived from a gastric protein ^[13]. Its proposed mechanisms emphasize VEGF-driven angiogenesis, growth hormone receptor upregulation, and direct tendon-fibroblast effects. TB-500 emphasizes actin-mediated cell migration and a different angiogenic pathway through the actin-binding motif. The thinking on forums: different mechanisms, complementary effects, better outcomes stacked than either alone.

The thinking is not wrong on its face. It's also not validated in any human study. Zero clinical trials have compared BPC-157, TB-500, or the combination head to head in any indication.

What the evidence base actually looks like:

BPC-157: broader animal literature, particularly in tendon and GI models, no completed registered human trials, the same compounding exclusions as TB-500, the same research-chemical regulatory posture.

TB-500: narrower animal literature for the fragment specifically, broader animal and limited Phase II data for the parent Tβ4 protein, same regulatory posture.

The honest read: the stack is community convention, not science. Some users report better outcomes than either alone. Whether that's pharmacology or expectation or placebo or the simple fact that anyone running a six-week injectable protocol is also probably resting, sleeping more, and eating better, nobody has separated. For a structured side-by-side, see our BPC-157 vs TB-500 comparison.

The "research use only" label on a TB-500 vial does specific legal work. It signals that the product has not been manufactured under pharmaceutical-grade conditions, has not been tested for human administration, and is being sold for in vitro and animal study only. It's the disclaimer that lets the vendor stay on the right side of FDA regulations governing unapproved drug marketing.

It's also the disclaimer that customers routinely ignore.

Worth understanding: a "research use only" product is not a pharmaceutical with a paperwork problem. It's a chemical compound from a supply chain that has no obligation to verify identity, purity, sterility, endotoxin level, or concentration. Some vendors do third-party test their products. Many don't. The COA (certificate of analysis) that accompanies a vial is generated by the vendor or a lab the vendor selected, with whatever rigor the vendor chose to apply.

If the product is going into a body, this matters more than the molecule itself.

The honest take, after stepping back from the file:

TB-500 is a real molecule with a real mechanism and a real preclinical signal. The biology is interesting enough that pharmaceutical interest in the parent Tβ4 protein produced legitimate Phase II clinical work over the last two decades. The translation from that work to the fragment being sold as a research chemical is mechanistically plausible and clinically unproven.

For adults considering it with eyes open, the relevant profile is narrow: a specific musculoskeletal injury where standard care has been exhausted, a clear understanding that this is an unregulated experiment with an unapproved compound, a vendor with third-party testing, sterile injection technique, and acceptance of both long-term unknowns and short-term contamination risk.

The contraindications are easier to enumerate. Anyone with a personal or family history of cancer. Anyone with active or suspected malignancy. Anyone subject to WADA or sport-organization testing. Anyone pregnant or attempting pregnancy. Anyone with real cardiovascular disease. Anyone using it for vague longevity reasons rather than a specific injury. Anyone who can't reliably source tested product. Anyone who wouldn't be comfortable explaining the decision to their physician.

What to watch for, if you're following this category over the coming years: any registered human trial specifically of TB-500 (not Tβ4) would shift the conversation considerably. So would any meaningful third-party testing program in the research-chemical supply chain. So would any regulatory action, positive or negative, from the FDA on Tβ4 itself. Until then, the file stays where it is: real biology, thin human evidence, an unregulated supply chain, and a community that has decided to act on the mechanism without waiting for the data.

That's not a verdict. It's the honest state of the file.