BPC-157 Before and After: The Rodent Timeline, the Thin Human Record, and What to Measure

Here is the problem with "BPC-157 before and after" stated plainly: almost all of the evidence is in rats. A 2026 review in Pharmaceutics counted the entire human record at fewer than 30 subjects across three uncontrolled pilot studies, none using a standardised pharmaceutical preparation. A scoping review in the American Journal of Sports Medicine the same year found 67% of the musculoskeletal literature was animal work and the human studies were "a handful", mostly without controls. There is no phase 2 trial, no validated human dose, and no human timeline.

So this page does something different from the others in the Bureau's results series. It sets out what the rodent tendon and gut data actually show day by day, because that is the only timeline that exists; it explains why rat days do not convert to human weeks; it lists what researchers running BPC-157 protocols report tracking, and how to record a baseline so that whatever happens can be judged; and it is specific about where a perceived result is far more likely to be time, rest or expectation than the peptide. The pharmacology is in the Bureau's BPC-157 guide.

What the evidence base actually contains

The tendon data come from a Zagreb group's rat studies. In the 2003 study, the right Achilles tendon was cut 5 mm above its insertion and BPC-157 was given intraperitoneally once daily at 10 µg/kg, 10 ng/kg or 10 pg/kg, starting 30 minutes after surgery, with assessments on days 1, 4, 7, 10 and 14. In the 2006 study the tendon was detached from the calcaneal bone and animals were assessed out to day 21. A 2011 study worked out part of the mechanism in cultured rat tendon fibroblasts: faster outgrowth from explants, better survival under oxidative stress, more migration, via the FAK-paxillin pathway.

The gut data are wider and older. BPC-157 was isolated from gastric juice, was in clinical trials for ulcerative colitis under Pliva as PL 14736 (the 2003 and 2006 papers describe it as "currently in clinical trials"; no results were ever published as a completed trial), and has an extensive rat literature on ulcers, colitis, fistulas and anastomoses summarised in the group's own reviews.

The human record: a 2021 retrospective chart review from a Florida clinic in which 16 knee-pain patients were phoned six to twelve months after an intra-articular injection and 11 of 12 who had BPC-157 alone reported "significant improvement", with no pain scale, no imaging and no control group; a two-subject pharmacokinetic pilot confirming a plasma half-life under 30 minutes; and small pilots in interstitial cystitis and intravenous administration described in the 2026 IJMS review. That is all.

Regulatory position, because it bears on what the "after" is worth: the FDA placed BPC-157 on its list of bulk substances that may present significant safety risks in September 2023, citing immunogenicity and impurity concerns; the current version of that page, dated April 2026, lists it as nominated but withdrawn. On 23 July 2026 the FDA's Pharmacy Compounding Advisory Committee voted 8 to 6 with one abstention to recommend adding it to the 503A bulks list, against the position of the agency's own scientists, whose briefing said the five available trials were "short in duration, small in sample size, and insufficient to establish safety or effectiveness". The vote is non-binding and rulemaking has not happened. The Bureau's FDA PCAC vote page follows that process.

The timeline: rat days against human weeks

Two things about the conversion. A rat's transected Achilles heals to functional integrity in two to three weeks with or without treatment, so the studies measure acceleration, not rescue, and the acceleration is measured in days. Second, the dosing in every rat tendon study was intraperitoneal, daily, at 10 µg/kg, which for a 70 kg human would be 700 µg by a route nobody uses; and the peptide's plasma half-life in the two-species ADME work and the human pilot was under 30 minutes. How that translates to a subcutaneous protocol is unknown, which the 2026 Pharmaceutics review calls a "pharmacokinetic-pharmacodynamic disconnect".

What to record before the first dose

Because no human trial defines an endpoint for BPC-157, the researcher has to define one, and the "before" has to be specific to the injury. The generic list the Bureau suggests:

• A numeric pain scale for a defined action, not "how does it feel". For an Achilles: pain out of 10 on the first ten steps in the morning, and pain on a single-leg hop. For an elbow: pain on a gripped 5 kg lift. For a shoulder: pain at 90 degrees of abduction. Same test, same time, three days running, averaged.
• A validated functional score for the site: VISA-A for the Achilles, VISA-P for the patellar tendon, PRTEE for the elbow, KOOS for the knee. These exist precisely because "it helped" is not a measurement, and the knee study that is BPC-157's best human evidence used none of them.
• A loaded capacity count: single-leg calf raises to failure, or pain-free grip strength on a dynamometer. Capacity is the thing tendons regain, and it is the hardest metric to fool with expectation.
• Imaging if the injury justifies it: a baseline ultrasound with tendon thickness in millimetres and a note of any neovascularisation, or an MRI. Without this there is no structural "after" at all.
• A load log: training volume, rehab exercises, rest days. If load drops by half the week the peptide starts, the peptide will get credit for the rest.
• For gut protocols: a bowel-habit and symptom diary scored daily, and, if a physician is involved, a standard laboratory marker of gut inflammation such as faecal calprotectin. The oral route is discussed on the oral BPC-157 page.
• Injection-site and systemic log: the FDA's stated concern is immunogenicity, so site reactions and anything resembling an allergic response are the adverse events worth writing down.

Realistic magnitude, and what is noise

The realistic magnitude in humans is unknown. That is not evasion; it is the finding of every 2026 review the Bureau read. The rat magnitude is large: full integrity by day 14 against a persisting defect in controls, and healing in the detachment model where controls did not heal at all. But the same group reports that BPC-157 heals nearly every tissue it is tested on in rats, from cornea to bowel to muscle, and results that uniform from a single laboratory are, in the words of the AJSM review, "promising but variable" until replicated elsewhere.

Noise, for a tendon, is enormous. Tendinopathy fluctuates week to week with load, sleep and weather. Rest alone improves pain within two to four weeks in most cases. Rehabilitation protocols produce measurable gains at six to twelve weeks. Placebo effects on pain are large and well documented. A researcher who starts BPC-157, drops training volume, does the rehab exercises and feels better at week four has four candidate explanations, three of which are not the peptide. The only honest way to shrink that list is a baseline capacity count and an unchanged load log.

For the gut, the same applies with a shorter clock: gastric symptoms respond to diet, alcohol and NSAID changes within days, which is exactly the window a protocol starts in.

What reverses on stopping

If BPC-157 accelerates tissue repair, as the rat data suggest, then a healed tendon stays healed: the peptide is not maintaining anything after the fact, and stopping should not undo structural repair. The rat studies stopped dosing at autopsy and never tested this, but it follows from the mechanism. What would be expected to reverse is any effect that depends on the peptide's continued presence, for which the candidates are its reported actions on nitric oxide signalling, blood vessel recruitment and pain modulation, all documented in rodents. A researcher whose pain returns within days of stopping is, on this reading, seeing the loss of a symptomatic effect, not a re-injury, and should treat that as evidence the underlying tissue had not changed.

For gut symptoms, the same logic: relief that disappears within a week of stopping was symptomatic. Nothing in the human record can confirm or refute either statement. The peptide cycles page covers how protocols are typically structured around this uncertainty.

Where people misattribute results

Rest as peptide. The most common one. Starting a compound is usually accompanied by taking the injury seriously, which means less load, and less load is the first-line treatment for most tendinopathy.

Rehab as peptide. Progressive loading has decades of controlled evidence for tendons. BPC-157 has none in humans. When both start the same week, the one with the evidence base is the likelier cause.

Time as peptide. Acute injuries heal. A grade 1 strain resolves in two to three weeks regardless. The rat data show acceleration of a process that completes on its own, and the human equivalent would be indistinguishable from ordinary healing without a control.

The knee study as evidence. It is a retrospective phone survey of 16 people with no pain scale, no imaging and no comparison group, published in an alternative-therapies journal. It is cited on nearly every vendor page. It cannot establish an effect size.

Stacks. BPC-157 is most often run with TB-500, and the Bureau's Wolverine stack and TB-500 versus BPC-157 pages explain what each is supposed to add. A result from the pair cannot be assigned to either.

Oral versus injected. The rat gut studies used oral and intraperitoneal routes; the tendon studies used intraperitoneal only. A researcher taking capsules for a tendon is outside even the animal data.

The FDA vote as approval. The July 2026 committee recommendation was advisory, was opposed by the agency's reviewers, and has not produced a rule. Nothing about BPC-157's evidence changed that day; only its politics did.

Sourcing and dosing notes

There is no validated human dose. Research protocols commonly describe 250 to 500 µg subcutaneously once or twice daily near the injury site, a range derived from scaling the rat 10 µg/kg figure rather than from any human study, and the BPC-157 dosage page explains where those numbers come from and what they do not tell you. The BPC-157 dosage calculator converts a vial size and water volume into syringe units for whatever protocol a researcher settles on.

On sourcing: the FDA's stated concern with this compound is impurities and characterisation of the active ingredient, which makes a per-batch certificate of analysis the one document that matters. The Bureau's top-scored source for BPC-157 at the time of writing is Apollo Peptide Sciences, which publishes the batch COA on the product page and ships from the US. Check price at Apollo Peptide Sciences The peptide testing page explains how to read the HPLC report before paying, and BPC-157 side effects lists what the human pilots and the FDA have and have not reported.

BPC-157 is an unapproved research compound with, as of September 2026, no completed controlled human trial. Everything above describes rodent studies and uncontrolled pilots; nothing here is medical advice, and the Bureau does not advise anyone on using research compounds.

Frequently Asked Questions

How long does BPC-157 take to work on a tendon?

In rats, days: treated animals had better functional scores within the first week after Achilles transection and full tendon integrity by day 14, with tendon-to-bone healing by day 21 in a detachment model. In humans, nobody knows. There is no controlled human tendon study, and a 2026 review counted fewer than 30 human subjects in the whole literature. Human tendon remodelling runs on months, so any change felt in the first three weeks is at least as likely to be rest, rehab or expectation as the peptide.

Is there any human evidence for BPC-157 before and after?

Very little. The most-cited paper is a 2021 retrospective survey in which 16 knee-pain patients were phoned six to twelve months after an intra-articular injection and 11 of 12 given BPC-157 alone reported significant improvement; it used no pain scale, no imaging and no control group. Beyond that are a two-subject pharmacokinetic pilot and small uncontrolled pilots in interstitial cystitis and intravenous use. The FDA's July 2026 briefing called the available trials short, small and insufficient to establish safety or effectiveness.

What should I measure before starting BPC-157?

A numeric pain score on a defined action, repeated three mornings and averaged; a validated functional score for the site such as VISA-A, VISA-P, PRTEE or KOOS; a loaded capacity count such as single-leg calf raises to failure or dynamometer grip; a baseline ultrasound or MRI if the injury warrants it; and a training and rehab load log. The load log is the one people skip, and it is the one that stops the peptide taking credit for a fortnight of rest.

Does BPC-157 keep working after you stop?

If it accelerates structural repair, as the rat studies suggest, a healed tendon should stay healed, because the peptide is not maintaining the tissue afterwards. What would be expected to fade is any effect that depends on its presence, such as the nitric oxide, blood-flow and pain-modulation actions reported in rodents. Pain that returns within days of stopping is therefore better read as the loss of a symptomatic effect than as re-injury. No human study has tested either possibility.

Why do the rat results not translate directly to people?

Three reasons. Rat tendons heal to function in two to three weeks with or without treatment, so the studies measure acceleration over days, not rescue over months. The rat dose was 10 µg/kg intraperitoneally every day, a route and scale no human protocol uses, and the peptide's plasma half-life is under 30 minutes in both species. And nearly all the tendon work comes from one laboratory, which a 2026 sports-medicine review describes as promising but unreplicated. The 2026 Pharmaceutics review calls the gap a pharmacokinetic-pharmacodynamic disconnect.

Did the FDA approve BPC-157 in 2026?

No. On 23 July 2026 the FDA's Pharmacy Compounding Advisory Committee voted 8 to 6 with one abstention to recommend adding BPC-157 to the list of substances compounding pharmacies may use. The vote is non-binding, the agency's own scientists opposed it, and the FDA must still decide whether to proceed through proposed and final rulemaking. BPC-157 remains an unapproved drug, and the FDA's compounding page, dated April 2026, still lists its concerns about immunogenicity and impurities.

Research use only. The compounds discussed are sold as research chemicals and are not approved for human use. Nothing here is medical advice.