How Long Do Peptides Take to Work? Honest Timelines by Category

All prices are all-in monthly: medication, physician review, refill management and shipping.

How to read that table, and which rows are trial endpoints

The middle column is the honest one and the one most pages leave out. "What patients commonly describe first" is not what the medication is doing, and not what a trial measured — it is whatever becomes noticeable first, and the order things become noticeable is rarely the order they happen in.

Two categories here have randomised human trials with published time courses. Semaglutide and tirzepatide have large phase 3 programs and approved labels specifying a titration schedule. Tesamorelin — the growth-hormone-releasing factor inside several of the blends — has a 26-week randomised endpoint in a named population. Everything else is mechanism plus what prescribers describe, and this page labels it that way each time.

What the animal literature actually times

The published work on BPC-157 is preclinical. There are no large randomised controlled trials of BPC-157 in humans. The animal studies are specific about their own timelines: a 2006 paper in the Journal of Orthopaedic Research examined Achilles tendon detachment in rats and reported accelerated reattachment and functional recovery, and work in the Journal of Applied Physiology in 2011 described tendon outgrowth, cell survival and cell migration in culture. Rat tendon does not heal on a human schedule, and reading a two-week rodent result as a two-week human result is the commonest mistake made with this literature.

TB-500 is a fragment of thymosin beta-4, described in Trends in Molecular Medicine in 2005 around actin sequestering and cell motility. Thymosin beta-4 does have human trial data in specific indications — a randomised placebo-controlled dose study in healthy volunteers in the Annals of the New York Academy of Sciences in 2010, and a phase 2 randomised trial in severe dry eye in Cornea in 2015 — but not for tendon or soft-tissue injury. The BPC-157 + TB-500 stack and the KLOW blend exist because the mechanisms are complementary, not because two peptides finish sooner than one.

Why rehabilitation is the actual rate limiter

For tendinopathy in particular, progressive loading under a physiotherapist has human trial evidence behind it that BPC-157 does not. A repair peptide supports a process driven by graded mechanical load, so someone doing the rehabilitation will describe a different timeline from someone who is only injecting — and the difference is not the peptide. This is also why the useful early sign is not "the pain is gone" but "the area is less reactive the day after I load it": a change in tolerance appears before a change in resting symptoms, and is only visible if you are loading the tissue. A problem present for a year is generally described as slower than a recent one, and one 28-day fill may not be the whole conversation.

GHK-Cu has the deepest research base of the repair peptides, including human work. A 1988 paper in FEBS Letters reported stimulation of collagen synthesis in cultured fibroblasts. The human studies are mostly topical and run on a months-long clock: work in Wound Repair and Regeneration in 1994 on topical GHK-Cu in patients with diabetes, and in Archives of Facial Plastic Surgery in 2006 on a topical copper tripeptide complex after CO2 laser resurfacing. Clinical data for injectable use specifically remains limited. The pattern matches the mechanism — collagen has to be synthesised, deposited and then remodelled, and remodelling is the slow step. Nothing in this category has a two-week answer.

Why photographs are the only honest measure

You see your face every day, which makes you the worst available judge of a gradual change in it. The standard advice from prescribers is a photograph in the same light, at the same distance, at the same time of day, at the start and at three months. Impressions at week six are not evidence in either direction. This applies equally to RADIANCE, ETERNAL and GLOW, all of which contain GHK-Cu.

Some of this catalog does not accumulate at all, and for those the question has a different shape entirely.

PT-141, and a number from an approved label

PT-141 is bremelanotide, a melanocortin receptor agonist. The approved product carrying that molecule is dosed as needed, at least 45 minutes before anticipated sexual activity, with no more than one dose in 24 hours and no more than eight doses a month recommended; the label also states that the duration of effect after each dose is unknown and the optimal administration window has not been fully characterized.

That is as specific a timing instruction as anything on this page, and it tells you what kind of medication PT-141 is: measured in minutes per dose and judged over a handful of uses rather than weeks. The randomised data on bremelanotide in premenopausal women — the dose-finding trial in Women's Health in 2016, the long-term analysis in Obstetrics & Gynecology in 2019 — was collected on that per-dose basis, as applies to PT-141 + Oxytocin.

NAD+, glutathione, Semax, Selank and DSIP

NAD+ and glutathione are cofactor and antioxidant repletion rather than signalling peptides. A pilot study in Frontiers in Aging Neuroscience in 2019 tracked the plasma and urine NAD+ metabolome during a six-hour intravenous infusion — the measurement window was the infusion itself. What people describe with NAD+ is same-session or next-day; what the literature measures is a blood level, not a sensation.

Semax and Selank are the most variable category here. The clinical literature on Selank in anxiety disorders, published in Zhurnal Nevrologii i Psikhiatrii in 2008 and 2014, ran on courses of weeks, while the mechanistic work — BDNF and trkB expression for Semax in Brain Research in 2006, GABAergic gene expression for Selank — is animal and cell research. Against that, many patients describe a same-day effect and many describe nothing at all; both reports are common and neither is wrong. If you are in the second group after several weeks, say so to your physician rather than escalating on your own. The nasal formats (Semax Nasal, Selank Nasal) are an alternative route, not a faster one.

DSIP has the oldest and most equivocal human literature in the sleep category — small insomnia trials in the Lancet in 1981 and European Neurology in 1986, less positive double-blind work in Neuropsychobiology in 1992, and a 2006 review in the Journal of Neurochemistry calling it an unresolved question. Expect the widest variation of any category here, and record it rather than remembering it.

One further effect belongs on that list. People start treatment when things are bad, and bad weeks are followed by average weeks whether or not anything was done, which flatters week two and makes week six look like a plateau — this hits sleep and pain reports hardest and appetite reports least.

Write down the single thing that would have to change for this to be worth continuing, before the first injection — picking it afterwards invites you to pick whatever happened to move. "Feeling better" is not measurable; "I can do a full session without the tendon being sore the next morning" is.

Then use the right instrument for the category. Appetite and weight: a four-week trend, not a daily weight. Growth-hormone axis: IGF-1 on a blood draw plus a written sleep and recovery note. Repair: tolerance to a specific load, plus your physiotherapist's own measures. Skin: photographs in fixed conditions. Intimacy: a count over several uses. Cognitive: whether someone who sees you daily notices anything, which is a harsher and better test than self-report. Bloodwork is not routinely required for most of this catalog and your physician decides; where it applies it is usually IGF-1, and metabolic and thyroid markers where the question is why a weight response is slower than expected. At-home blood testing covers those without a lab visit.

Refills run on a 28-day rhythm because that is roughly what a vial covers at typical directions, and treatment here is long-term rather than a fixed-length course. That rhythm is useful as a checkpoint for a separate reason: four weeks is the shortest interval over which a weight trend, a sleep record or a load-tolerance note says anything at all.

So the practical structure is one fill, one written record, one honest look at it, and a message to your physician if the record says nothing has moved — not a decision made on day nine. What happens when you stop taking peptides covers the other side of the question.

• A side effect that is not settling, or one that is getting worse.
• Nothing at all by the reassessment point your physician described.
• A new symptom you did not have before starting.
• An injection site that stays red, hot or painful.
• Anything that worries you enough that you are searching for it online.

And three things not to do while waiting: increasing your own dose, adding something else at the same time so neither change can be interpreted, or stopping a GLP-1 abruptly without telling anyone. All three make the next clinical decision harder. What stopping actually does is covered in what happens when you stop taking peptides.

The middle one deserves emphasis. Start a repair peptide and a GH-axis peptide in the same week and you have two variables and one observation, with no way to attribute anything. That is the most common reason a physician cannot answer whether something is working — not because the answer is unknowable, but because the experiment was destroyed.

None of the compounded medications on this page are FDA approved as finished drug products. Compounded medications are prepared by a licensed pharmacy pursuant to a prescription rather than approved as manufactured products, and that holds regardless of how much research exists behind the molecule. Are peptides FDA approved covers the distinction properly.

The FDA's list of bulk drug substances that may present significant safety risks under the 503A interim policy — commonly called Category 2 — is current as of 22 April 2026 and no longer includes BPC-157, thymosin beta-4, GHK-Cu, epitalon, DSIP, selank or semax; the page records those nominations as withdrawn by the nominators. The FDA's parent page on 503A bulk drug substances, current as of 14 May 2026, sets out how Categories 1, 2 and 3 work. The Pharmacy Compounding Advisory Committee met on 23–24 July 2026, and FDA determinations following that meeting are issued by the agency rather than by the committee.

Semaglutide and tirzepatide are approved as branded finished products, and the titration schedules quoted above come from those approved labels. The compounded preparations share the active ingredient and not the approval.

Sometimes nothing has happened because nothing was going to, and the honest version of this page has to say so.

• The problem is structural. A complete rupture, a displaced fracture or a significant meniscal tear is a surgical question. No repair peptide changes that, and waiting longer only costs time.
• The problem is a deficiency. Fatigue from low B12 is a question for methylcobalamin, not a growth-hormone secretagogue. An untreated thyroid problem responds to neither.
• The category is wrong for the goal. If the goal is visceral fat specifically, the 26-week evidence sits with tesamorelin and the blends containing it rather than with a repair peptide. If the goal is systemic inflammation rather than a local injury, KPV or FORTIFY is the more relevant conversation.
• The expectation was never achievable on that schedule. Expecting a skin change at four weeks, or a body-composition change at three, is not a treatment failure. It is a calendar problem.

Your physician decides which of these applies and whether a different product is appropriate. A consultation does not guarantee a prescription. How do I know if peptides are working covers the measurement side of the same question.

For the GLP-1 medications, the time course of appetite and weight change is documented in large randomised trials of the branded products. The compounded preparations share the active ingredient but not the approval, and no trial has been run on a compounded preparation.

For everything else in this catalogue, the timelines above come from mechanism and from what patients report to prescribers, not from trial endpoints. The published research on BPC-157, TB-500, KPV, MOTS-c and most of the rest is preclinical: animal and cell studies. GHK-Cu has the deepest literature of the repair peptides, including human skin work, though clinical data for injectable use remains limited.

Tesamorelin is the notable exception in the growth-hormone-axis category, with a 26-week randomised endpoint in a named population, and the GHRH(1-29) literature behind sermorelin includes controlled human studies with published time courses for IGF-1. Thymosin beta-4, the parent of TB-500, has human trial data in indications other than tendon injury. Naming those accurately matters as much as being honest about what is missing elsewhere.

That means the honest version of this page is a set of expectations rather than a schedule. Compounded medications are not FDA approved, response varies between people, and a provider quoting you a number of days until a result is telling you something about the provider.