If your knees or tendons hurt, you have likely seen peptides pitched as a fix. The word covers very different things. Some are food supplements. Others are injectable drugs that were never approved.
So take them one at a time. For each, the question is the same: has anyone tested it for joint pain in people, against a placebo? The broader record for the injectables is in BPC-157 human evidence.
| Peptide | Human joint studies | Control group? |
|---|---|---|
| Collagen peptides (oral) | Several randomized trials | Yes, placebo |
| BPC-157 (injected) | One chart review, 16 patients | No |
| TB-500 / thymosin beta-4 | None published | — |
| GHK-Cu (injected) | None published | — |
Collagen peptides
Collagen peptides are small pieces of collagen, taken by mouth as a powder or capsule. They have the most human data of anything on this list.
A 2023 meta-analysis pooled 4 trials with 507 people with knee osteoarthritis [1]. Pain fell more on collagen than on placebo (standardized mean difference −0.58, 95% CI −0.98 to −0.18).
Here is the catch. The authors rated all 4 trials at high risk of bias. Adverse events did not differ clearly from placebo, but the evidence on that was very low quality.
One newer trial gave 80 adults with mild knee osteoarthritis 3,000 mg a day for 180 days [2]. Pain scores fell more than on placebo (p = 0.006), and no adverse events occurred. Two authors had industry affiliations.
Collagen has also been tested in younger people with sore knees from exercise. In 180 active adults aged 18 to 30, 5 g a day for 12 weeks cut exercise pain scores by 21.9 mm, against 15.6 mm on placebo (p = 0.024) [3].
That is a real but modest gap. A wider review of joint supplements found short-term pain effects but none that were clinically important over the long term [4].
BPC-157
BPC-157 is widely marketed for joint and tendon injuries. Its entire human joint record is one retrospective chart review from a single clinic [5].
The clinic phoned 16 patients who had BPC-157 injected into the knee, alone or with thymosin beta-4. Of the 12 given BPC-157 alone, 11 said their pain improved [5].
There was no placebo group, no standard pain scale and no imaging. A 2026 review for orthopedic surgeons called its flaws significant [6]. The side-effect picture is in BPC-157 side effects, and the one study that tested it with TB-500 is in BPC-157 and TB-500 together.
TB-500 and thymosin beta-4
TB-500 is a fragment of thymosin beta-4. The same 2026 review found repair effects in animals, but no human orthopedic data [6]. The FDA says it has found no human exposure data for TB-500 at all [7]. More detail is in TB-500 evidence.
GHK-Cu
GHK-Cu is mostly used on skin. The 2026 orthopedic review found no clinical data supporting it for musculoskeletal conditions [6]. Its safety record by route is in GHK-Cu side effects.
What this means for you
If you want the option with the most human testing, it is oral collagen. The benefit is modest and the trials are imperfect.
The injectable peptides are a different bet. Nobody has run a controlled human trial of them for joint pain. Their legal status is covered in are peptides legal.
For recovery options with human data behind them, start at the recovery goal page.