Most healing-peptide discussion starts and ends with BPC-157 and TB-500. That pairing covers two early bottlenecks, blood flow and repair-cell movement, and for straightforward soft-tissue cases that is often enough. The more useful question is what to look at when an injury stalls anyway, because a stall usually means one specific step in the repair process is blocked. This sorts the common compounds by the bottleneck each one targets, with evidence level and reported research-context dosing in tables you can skim.
One thing first, because it shapes how to read everything below. No human randomized controlled trials exist for any multi-peptide healing combination. Individual compounds have preclinical support, much of it in rodents, and a few have early human data. Every stacking rationale is mechanistic, built from how the compounds behave in isolation rather than from controlled outcome trials.
Why injuries stall: the five repair bottlenecks
An injury that has not improved after four to six weeks is usually blocked at a specific step rather than just healing slowly. Five bottlenecks explain most soft-tissue stalls.
Bottleneck What's happening Signs Mapped compound Perfusion Tissue starved of blood flow, oxygen, immune cells Stays cold or pale, swelling persists for weeks, PT gains don't hold BPC-157 (angiogenesis) Cell migration Repair cells can't move through and organize Less cold but stiff and locked, ROM improves then stalls, fascia feels stuck TB-500 Cellular energy Compromised mitochondria can't fund repair Healing starts then plateaus, setbacks with modest activity NAD+, then SS-31 Collagen quality New tissue forms weak, disorganized fibers Re-injury at the same site, tissue feels loose or unstable GHK-Cu Inflammation gating Inflammation switch won't shut off Better after activity then worse 2 days later, pain worse with rest KPV (NF-kB)
Two more show up later in recovery rather than upfront. Disrupted deep sleep closes the window when GH-driven overnight repair happens (waking unrefreshed, recovery stalls despite good rehab compliance): DSIP is associated with restoring deep-sleep architecture, Selank with the stress and anxiety that fragment sleep. Blunted or mistimed GH release is the other (cortisol stays elevated overnight, GH pulses weak or mistimed): tesamorelin or sermorelin map here. These overlap often, since a tendon injury with poor blood supply tends to develop chronic inflammation because weak circulation prevents immune-cell clearance.
What the evidence shows, by tier
The compounds are not equally supported, and the gap is wide. This is the honest split.
Tier 1, actual human data
BPC-157 has the broadest preclinical base. A 2025 systematic review screened 544 articles and included 36 studies: 35 preclinical, one clinical. The animal data across tendon, ligament, muscle, and bone is consistent and reproducible, showing improved biomechanics and faster functional recovery, and it can counteract corticosteroid-impaired healing in preclinical models. The review authors still caution against clinical use, specifically because the human evidence is so thin and in-human safety is unknown. The intra-articular knee case data that gets cited a lot, a series reporting symptom reduction in over 90% of 17 patients, is part of that single clinical study, and the review's caution applies to it directly.
ARA-290 (cibinetide) is one of the few with real Phase 2 human data. In a 64-patient trial in sarcoidosis-associated small fiber neuropathy, the 4 mg dose significantly increased markers of nerve fiber regeneration. The precise read matters: the structural nerve-regrowth signal was real, but pain reduction did not separate from placebo in that dose-ranging trial. So the defensible claim is nerve-fiber regeneration evidence, not established pain relief.
Tier 2, mechanistic plus preclinical, no injury RCTs
NAD+, GHK-Cu, KPV, SS-31, Selank, DSIP. Each has a mechanistic rationale and varying preclinical support, but none has human RCT data for injury healing specifically. These are reasoned additions for a defined bottleneck, not validated treatments.
Tier 3, thinnest, single-model or non-injury data
VIP (disc data from a single mouse model, very short ~1 min half-life), Cartalax (mostly Russian-language bioregulator literature, no injury-recovery data), relaxin-2 (reverses joint capsule scarring in mouse models, entirely preclinical).
Reported research-context dosing
These come from practitioner and community literature, not human dose-finding trials. Listed for reference, not as protocol.
Compound Reported range Notes BPC-157 250–500 mcg daily Oral alt: PDA, same 15-AA sequence, arginate salt, ~500–1000 mcg/day, systemic not local TB-500 2–4 mg 2x/week Loading block then weekly taper; field/equine-derived ARA-290 2–4 mg daily Systemic injection, no need to place near nerve NAD+ 50–250 mg, few x/week Funds the energy pool the other compounds spend GHK-Cu 1–1.5 mg, few x/week Collagen remodeling KPV 200–500 mcg daily Inflammation gating SS-31 5–10 mg Cardiolipin binding; consider when healing stalls 4+ wks on core Selank 250–500 mcg SubQ or intranasal (intranasal bypasses liver) DSIP 100–300 mcg Extends deep-sleep window VIP 100–200 mcg daily Sometimes split; ~1 min half-life Tesamorelin 1–2 mg Cycled ~8–12 wks on, 4 off in practitioner lit Sermorelin 200–500 mcg Same cycling pattern
TB-500 product check: verify identity against a certificate of analysis. The fragment sits near 859 Da, full-length TB-4 near 4,900 Da, and they are not interchangeable. TB-4 can access pathways the fragment cannot, including Ac-SDKP anti-fibrotic biology, and most underlying research uses TB-4 (~50 preclinical studies plus Phase 1 human safety data).
Matching the add-on to the tissue
The core covers blood flow, migration, and energy. What gets layered on depends on the tissue.
Injury type Add Why Connective tissue / collagen quality GHK-Cu Tissue feels unstable, re-injury at same site Inflammation cycling KPV Swelling returns after activity Disc / spinal degeneration VIP Only peptide with direct disc data Nerve pain / neuropathy ARA-290 Phase 2 nerve-regeneration data Disc injury, energy angle SS-31 Cuts inflammation-induced disc-cell death via mitochondrial scavenging
The strongest preclinical case sits in tendon, ligament, and fascia, where BPC-157 promotes tendon repair-cell outgrowth through FAK-paxillin signaling, rat Achilles studies show accelerated healing with improved biomechanical strength, and TB-4-treated tendons showed more uniform fiber bundles with increased collagen fibril diameter versus controls. Plantar fascia is structurally similar but has no direct peptide data, so the same mechanisms are plausible but unvalidated there. Focal cartilage defects are a different problem: the tissue is avascular and matrix-limited, and intra-articular BPC-157 in that context is clinician-only, not an at-home translation.
Supporting factors that aren't optional: collagen peptide supplementation at 5–15 g/day, 30–45 min before tendon-loading work, with vitamin C as a required cross-linking cofactor and glycine as the most abundant amino acid in collagen. Progressive mechanical loading is non-negotiable, since collagen fibers align along lines of stress and without controlled loading new tissue forms as disorganized scar. Peptides and rehab are synergistic; neither replaces the other.
The GH-axis caveat worth knowing
GH secretagogues come up for overnight recovery, but the injury context changes the calculus. Growth hormone can cause mild joint swelling, and in a joint that already has inflammation, cartilage damage, or impingement, that extra fluid can worsen symptoms rather than help.
The clearest human data is a randomized trial of recombinant growth hormone after arthroscopic rotator cuff repair:
Group n Healing failure rate rhGH 4 mg 26 30.8% rhGH 8 mg 24 16.7% Control 26 34.6%
All differences non-significant (p > .05), with no significant difference in healing, range of motion, or pain. Added GH did not improve the outcome. Ligament-driven injuries appear to tolerate GH-axis support better than cartilage or joint-surface problems, so more GH signaling is not automatically better, and anyone adding a secretagogue for overnight recovery has reason to watch for increased joint stiffness or swelling in the first couple of weeks.
Timeline as commonly described
Observational, from practitioner and community sources, not trial-derived.
Window What's described Weeks 1–2 Reduced pain, improved mobility Weeks 2–4 Active remodeling, new collagen deposition Weeks 4–8 Structural maturation under load Weeks 8–12 Full effects for chronic injuries
Acute injuries are generally framed at 4–8 weeks, chronic at 8–12, with old injuries running longer because of established scar tissue, adapted blood supply, and mitochondrial depletion. Week 8 is the common reassessment point: if progress has stalled, that's where collagen quality, sleep, or hormonal timing get evaluated as the next bottleneck.
Regulatory and testing status
None of these is an FDA-approved injury treatment. In April 2026 the FDA removed BPC-157, TB-500, KPV, and injectable GHK-Cu, among others, from the Category 2 restricted compounding list, following withdrawal of their nominations. That removal did not move them to the authorized list, which leaves them in a gray zone, neither explicitly prohibited nor permitted for compounding, with a Pharmacy Compounding Advisory Committee review scheduled for late July 2026. That is access context, not a therapeutic verdict. For tested athletes, BPC-157 and TB-500 are prohibited under WADA rules and have been banned across the NFL, UFC, and NCAA since 2022.
The takeaway
The framework earns its keep as a way to think about why a specific injury is stuck: find the bottleneck, then look at which compound is mechanistically matched to it. What it cannot do is promise outcomes. The mechanistic logic is reasonable and the preclinical signal for the core compounds is real, but the human outcome data runs from thin to absent depending on the compound, and no combination has ever been tested as a combination.