How BPC-157 is injected, from the record: the routes trials register, the one knee-injection chart review, an IV pilot in two people, site claims checked.
"How to inject BPC-157" is asked as if the answer were in a label. It is not: no BPC-157 product is approved anywhere, so no manufacturer has written the site, needle, volume or schedule that every approved injectable carries. What exists is a short human record (two registered trials, one chart review, one two-person infusion pilot), an animal literature that used routes no person would, and a large body of vendor and forum practice. This page reports each, dated 2026-10-06, and keeps them apart. It describes how injection is reported to be done; it instructs no one.
Reported amounts, the regulatory history and the per-vial arithmetic are on the BPC-157 dosage page; the oral route, which the earliest human trial used, is on oral BPC-157. The general subcutaneous technique from approved labels and nursing texts, which is where the procedure for any peptide comes from, is on how peptides are injected. The compound itself is covered in Medibact's BPC-157 guide. The same question for its usual partner, where the two rat studies gave the compound into the abdomen rather than at the injury, is answered on the TB-500 injection page.
What the human record says about routes
| Record |
Route |
Dose stated |
Site stated |
People |
| NCT07803250, University of Arkansas (not yet recruiting) |
Subcutaneous, self-administered, daily for 90 days |
No |
No |
30 planned |
| NCT07437547, Hudson Biotech (recruiting) |
Subcutaneous, by study staff, daily for 14 days |
No |
No |
120 planned |
| Lee and Padgett 2021, chart review |
Intra-articular (knee) |
No |
Knee joint |
17 (16 followed up) |
| Lee and Burgess 2025, pilot |
Intravenous infusion over 1 h |
10 mg, then 20 mg |
— |
2 |
| NCT02637284, PharmaCotherapia (status unknown) |
Oral tablets |
1 mg |
— |
42 planned |
Two things stand out. Both registered injection trials chose subcutaneous injection, the route vendor pages describe, and neither published the dose or the site. And the two records that did put BPC-157 into people at stated amounts or sites both come from Florida clinics with the same first author (E. Lee), with no control group in either. The Hudson Biotech record is one of eight that sponsor posted in early 2026 at one hospital in Shenzhen, one of which reproduces the title and enrolment of a completed Lilly trial (details on the dosage page); the registry lists records, it does not verify them. The Arkansas record has its own inconsistency: one passage says the sham group receives no injections, the arm description says it injects saline daily for 90 days.
The "inject near the injury" claim
Vendor, clinic and forum pages commonly describe injecting subcutaneously close to an injured tendon or joint, on the reasoning that the peptide acts locally. That reasoning is not what the animal studies show. In the rat ligament study cited most often for healing (Cerovecki et al., 2010), the injured knee ligament healed better when BPC-157 was injected into the abdominal cavity at 10 µg/kg or 10 ng/kg, when a cream was applied over the injury, and when it was given in drinking water: three routes, one of them nowhere near the injury, with similar results. A 2026 review from the same research group describes BPC-157 as acting "given systemically or locally" (Matek et al., Pharmaceuticals, PMID 41754849). No human study compared a site near an injury with any other site.
So the human question of "where" has no tested answer, and the animal record, taken at face value, suggests the site may not matter. Both statements are about absence of evidence, not evidence of equivalence.
Meniscus, back pain and other site searches
Searches name a body part: meniscus, back, shoulder, knee. PubMed (searched 2026-10-06) holds no BPC-157 study of back pain or the spine. Meniscus tears appear in exactly one human paper, the 2021 knee chart review, which listed them among many causes of knee pain, grouped every cause together, and gave no dose; a 2026 orthopaedic review called that study's methods significantly flawed and concluded that "information regarding the indications, dosing, frequency, and duration of treatment remains unknown" (Mayfield et al., American Journal of Sports Medicine, PMID 41476424). An injection into a joint is a clinical procedure; nothing about it can be read off a vendor chart. The registered Arkansas trial (rotator cuff) and Hudson trial (hamstring) inject subcutaneously, not into the shoulder or the muscle.
How the general technique is described
Because BPC-157 has no label, descriptions of its injection borrow the subcutaneous technique of approved peptide labels and nursing texts: a short insulin-syringe needle into the fatty layer, commonly of the abdomen or thigh, with sites rotated. Those sources, and what each one says, are set out on how peptides are injected; the syringe scale is on insulin syringe units. For approved peptides, the labels state that exposure is similar across the abdomen, thigh and upper arm. No equivalent measurement exists for BPC-157 at any site.
The amounts, as syringe units
Units on a U-100 syringe = mg ÷ mg/mL × 100. The figures in circulation are 250 to 500 mcg (vendor and clinic pages, as recorded on this site's BPC-157 page, 2026-09-18); research-market vials are labelled 5 mg and 10 mg.
| Amount |
2.5 mg/mL |
5 mg/mL |
10 mg/mL |
| 250 mcg |
10 units |
5 units |
2.5 units |
| 500 mcg |
20 units |
10 units |
5 units |
Where the concentrations come from: 2.5 mg/mL is a 5 mg vial with 2 mL; 5 mg/mL a 5 mg vial with 1 mL or a 10 mg vial with 2 mL; 10 mg/mL a 10 mg vial with 1 mL. The per-vial arithmetic is on the BPC-157 reconstitution page and the BPC-157 calculator. The two-person infusion pilot's 20 mg would be 800 units at 2.5 mg/mL, which is why it was given in 250 mL of saline rather than by syringe.
What is not established
Not established by any source read for this page: a human dose of injected BPC-157 supported by a controlled trial, any effect of injection site on outcome, any study of BPC-157 for back pain, and any human pharmacokinetic data after subcutaneous injection. The US position, including the 2026 compounding committee vote, is on the dosage page. Nothing here is a suggested amount, route or site; decisions about them belong with a clinician.
Sources and dates
Opened 2026-10-06: ClinicalTrials.gov API v2 (intervention search for BPC-157; records NCT07803250 and NCT07437547, arms and interventions); PubMed abstracts via NCBI E-utilities for Lee and Padgett 2021 (PMID 34324435), Lee and Burgess 2025 (PMID 40131143), Cerovecki et al. 2010 (PMID 20225319), Matek et al. 2026 (PMID 41754849) and Mayfield et al. 2026 (PMID 41476424); PubMed searches for BPC-157 with back pain, lumbar, intervertebral, meniscus and subcutaneous. The DailyMed result, the Hudson Biotech registry pattern, NCT02637284 and the figures in circulation are as recorded on this site's BPC-157 dosage page (sources read 2026-09-18). Unit arithmetic by this site. Corrections go to the contact page.
