Guide

Peptides for longevity: what the evidence actually shows

8 min read · last reviewed 22 August 2026

Peptides are the fastest-growing corner of longevity marketing, sold with an aura of clinical precision — exact doses, injection protocols, stacking schedules. The precision is real; the evidence behind it mostly is not. This guide walks the best-known longevity peptides in order of how much human evidence actually exists, which turns out to be the reverse of how much excitement surrounds them.

First, the regulatory reality

One distinction sorts the entire category: collagen peptides are a food supplement you can buy in a supermarket, while almost every other longevity peptide — BPC-157, epithalon, injectable GHK-Cu — is an unlicensed compound sold online 'for research use'. That label is not a formality. It means no medicines regulator has assessed the product, no manufacturing standard applies to it, and nobody verifies that the vial contains what the website says at the purity an injection demands.

Nothing on this page is a recommendation to buy any of them. The evidence pages exist because people are buying them anyway, and deserve to know what is actually established.

Collagen: the one with real trials

Collagen is the only peptide here with a replicated randomised-trial literature. A 2021 meta-analysis of nineteen trials supports improvements in instrument-measured skin elasticity and hydration; single trials support benefits for activity-related joint pain and for resistance-training gains in elderly men. Two caveats travel with all of it: the outcomes are narrow measurements rather than health outcomes, and most trials are funded by collagen manufacturers. Nothing connects collagen to ageing itself.

Collagen peptides evidence →

BPC-157: an animal literature wearing a lab coat

BPC-157 has the most confident dosing protocols on the internet and the least human evidence to support them. Rodent studies genuinely and repeatedly show accelerated tendon, muscle and gut healing. But a 2025 systematic review of its use in sports medicine found that no randomised human trial has ever been published — not one, on any endpoint. Every human protocol circulating online is extrapolation from rats, and there are no controlled human safety data for the injections people are taking.

BPC-157 evidence →

GHK-Cu and epithalon: biology without trials

GHK-Cu has attractive mechanistic biology — it modulates genes involved in skin repair, which is why cosmetics use it — but no randomised human trial behind its regenerative claims, and no human safety data at all for the injectable use now sold online.

Epithalon is the strangest case: cell-culture work supports telomere lengthening, and its human lifespan claim rests on a Soviet-era programme run by the peptide's own developers — non-randomised, and never independently replicated in fifty years. Worth knowing: telomerase activation, the proposed mechanism, is also the trick most cancers rely on. Our rubric grades what exists; the factor panels on the evidence page show exactly why the grades should be read sceptically.

GHK-Cu evidence →Epithalon evidence →

How to think about the next peptide you hear about

The pattern generalises. Ask three questions in order. Is it licensed — if not, purity and identity are unverified. Has any randomised human trial tested the claim you care about — for most peptides the answer is no, and the confident protocols are rodent extrapolation. And who ran the supporting studies — a literature that never leaves its originating lab is a warning sign the rubric can see, and so can you.

Licensed food supplement with replicated trials: collagen — narrow, cosmetic-functional claims hold up.
Unlicensed, rodent-only: BPC-157 — genuinely promising biology, zero human trials.
Unlicensed, mechanism-only: GHK-Cu injectables, epithalon — claims run far ahead of any human data.

Why animal results don't transfer →