BPC-157 adverse event reports to FDA are climbing. They cannot show what caused what.

Why this is in the news

On October 13, 2026, KFF Health News published accounts from people who say BPC-157 hurt them, including persistent fatigue and anxiety after nasal use and severe migraine with facial numbness after a BPC-157 and TB-4 stack.[6] These are individual accounts, not verified cases. The article also described two hospitalizations from FDA's adverse event database: a 36-year-old woman and a 48-year-old man with a life-threatening condition.[6]

The timing matters because FDA has sent an "Interim Access Policy for Compounding Using Certain Peptides" to the White House for review; on the list run October 16, 2026 it was still under review, with no text public.[9] Our October 16 note covers that filing, and our July note covers BPC-157's small human studies. This note is about the safety record.

FDA's own count

Three reports through December 2025

For its July 2026 advisory committee, FDA's Office of Surveillance and Epidemiology searched FAERS, its adverse event reporting database, for BPC-157 through December 4, 2025. The search retrieved three reports, all involving injectable BPC-157.[1] A 55-year-old woman had nine days of redness and swelling at the injection site while also injecting compounded thymosin. A 28-year-old man developed shortness of breath after subcutaneous BPC-157 acetate and went to an emergency room. A 40-year-old woman developed diffuse skin and gum darkening after a "research purposes only" product containing BPC-157 and TB500, with the same reaction on re-use.[1]

FDA judged each case hard to interpret: the thymosin co-injection, missing details, and a two-peptide product meant it was "not possible to assess" BPC-157's role in the third.[1] A separate search of FDA's food and supplement complaint system through December 3, 2025 found three more cases, all complicated by multi-ingredient supplements, other medicines or missing details, and FDA found no published BPC-157 case reports.[1]

Checked Oct. 17, 2026

What FDA's public data shows now

pepmg queried openFDA, FDA's public copy of FAERS, on October 17, 2026; its data was last updated July 30, 2026.[2] A search for drug names beginning "BPC" returned 30 reports. Ten were other or ambiguous products, mostly vitamin preparations whose names carry "BPC." Twenty name BPC-157; eight list it as a suspect product, and twelve list it only as something else the person was taking.[2] The eight include FDA's three.

Across the eight, four are coded as hospitalizations, three as life-threatening, and none as a death; all are coded serious.[2] Six came from consumers, one from a health professional who was not a physician or pharmacist, and one has no reporter type.[2] Five also name other suspect products, including Sermorelin , CJC-1295 and GLP1 S , and six carry coding for contamination, sterility, labeling or other product problems.[2] Two describe allergic-type reactions. The January and March reports match the ages, sexes and seriousness KFF described, but openFDA does not let us confirm they are the same cases.[2][6]

Reuters reported in September that FDA's Adverse Event Monitoring System, formerly FAERS, showed 10 reports since 2021, all serious, eight filed in 2026.[3] That does not match either of our counts. It may count reports differently or include data newer than openFDA's July update; pepmg could not reproduce it.

What reports can't do

A count of reports is not a count of harms

FDA's briefing is direct about this. There is "no certainty that the reported adverse event was due to the suspect product," FDA does not receive every event, and "FDA cannot make definitive conclusions regarding the safety of BPC-157 based on FAERS data alone."[1] FDA also notes that 503A compounders generally do not report adverse events to it.[1] openFDA's own notice says to assume its results are unvalidated.[2]

Two forces push in opposite directions. Under-reporting means the real number of problems is probably higher. Publicity means reports can rise because more people know to file them, not because more harm is happening. With no count of how many people use BPC-157, these reports cannot produce a rate.

Who is using it

A health-records preprint: 1,039 documented users

The closest thing to a denominator is a preprint posted September 7, 2026 and not peer reviewed. Researchers used a large language model, with physician validation, to search de-identified clinical notes from a U.S. federated health-records network. Of 1,536 patients whose notes mentioned BPC-157, 1,039 had documented use, and quarterly new confirmed users rose 33-fold between 2020 and 2026.[5] Reuters attributed the study to the analytics firm nference, with Venky Soundararajan as its leader.[3]

Among users, 95.9% of those with recorded race were White, and the share of new users who were men rose from 57% to 70% over the period. Pain was the most common documented reason, at 33% of users, and about half also used other drugs, most often testosterone.[5] Adverse events were noted in few records, each type in under 1.5% of users: neuropsychiatric 1.4%, injection-site hypersensitivity 1.4% and gastrointestinal 1.3%. The authors say these were likely under-reported.[5] Improvement was documented for 79% of the 354 users with a recorded response direction, but direction was missing for 65.9%, and the authors say the data cannot establish efficacy or separate BPC-157 from co-treatments, placebo or reporting bias.[5]

pepmg could not open the full preprint; these figures come from its abstract as recorded by Crossref.[4][5]

The trial record

Thirty people in three poor-quality studies

A systematic review published October 8, 2026 searched Embase, PubMed, Cochrane and ClinicalTrials.gov through June 1, 2026 for human BPC-157 studies; it focused on orthopaedic and regenerative use but accepted any indication. It found three: a retrospective knee-injection study of 16 people, a retrospective bladder-instillation study of 12, and an intravenous safety study of 2.[7] All reported no treatment-related adverse events, but the reviewers rated the overall certainty of evidence very low.[7] FDA's briefing counted five human studies, adding two rectal-enema studies, one in 24 healthy volunteers and one in people with ulcerative colitis; it found no serious adverse events reported but said safety monitoring in most was unclear.[1] Studies this small and short cannot show whether uncommon reactions happen.

FDA's compounding safety page lists BPC-157 among substances whose nominations were withdrawn and says the agency "lacks sufficient information to know whether the drug would cause harm when administered to humans."[8] BPC-157 is not an FDA-approved drug.[3]

Questions people are asking