Half-life reference.
Elimination half-lives taken from FDA prescribing information and peer-reviewed pharmacokinetic studies, with the source linked on every row. 24 compounds have a published human value we could trace to a primary source. 18 are listed precisely because a number circulates for them and no reliable human data stands behind it.
Nothing in this table tells you when to take anything, how often, or for how long. Labelled dosing intervals come from clinical trials that measured outcomes and tolerability, not from arithmetic on a half-life. This is a reference for understanding pharmacokinetics, not a timing tool.
42 of 42 entries
| Published half-life | Route | Notes | Source | |
|---|---|---|---|---|
| 5-Amino-1MQ | No reliable published human pharmacokinetic data | No human study published | Mouse plasma and tissue pharmacokinetics after intravenous, oral and subcutaneous dosing. No human data. | Babula et al., Diabetes Obes Metab 2024 (PMID 39161060)Animal data only |
| ACE-031ramatercept · ActRIIB-Fc | 10 to 15 days | Subcutaneous | Single ascending dose study in 48 healthy postmenopausal women, 0.02 to 3 mg/kg. Development of ACE-031 was later discontinued on safety grounds. | Attie et al., Muscle Nerve 2013 (PMID 23169607)Peer-reviewed |
| AfamelanotideMelanotan I · Scenesse | Approximately 15 hours (apparent) | Subcutaneous controlled-release implant | Apparent and release-rate-limited, with a median Tmax of 36 hours. The implant's clinical duration of about two months is a property of the delivery system and is unrelated to this figure. | Scenesse label, DailyMedFDA label |
| AOD-9604 | No reliable published human pharmacokinetic data | Not published | Phase 2 obesity trials were run by the original developer, but no pharmacokinetic paper for AOD-9604 is indexed and no half-life was reported in the human trial literature. | Wilding, Curr Opin Investig Drugs 2004 (PMID 15134286)No human PK |
| BPC-157body protection compound | No reliable published human pharmacokinetic data | No human study published | Rats and beagle dogs only. Mean elimination half-life was 15.2 minutes after intravenous dosing in rats. Rodent clearance does not scale reliably to humans for peptides, so this is not a substitute for a human value. | He et al., Front Pharmacol 2022 (PMC9794587)Animal data only |
| Cagrilintideamylin analog | 159 to 195 hours | Subcutaneous | Range across the 0.16 to 4.5 mg dose cohorts, median Tmax 24 to 72 hours. The same trial independently measured semaglutide at 145 to 165 hours. | Lau et al., Lancet 2021 (PMID 33894838)Peer-reviewed |
| CJC-1295 | No reliable published human pharmacokinetic data | Not established for the non-DAC form | The published 5.8 to 8.1 day figure belongs to CJC-1295 with DAC only. The DAC is what binds albumin and produces that half-life, so it cannot be transferred to material sold without it. No controlled human study of the non-DAC form exists. | Teichman et al., J Clin Endocrinol Metab 2006 (PMID 16352683)No human PK |
| CJC-1295 DAC | 5.8 to 8.1 days | Subcutaneous | Healthy adults aged 21 to 61. The drug affinity complex binding to albumin is what produces this half-life, and it is the reason the DAC and non-DAC forms cannot share a number. | Teichman et al., J Clin Endocrinol Metab 2006 (PMID 16352683)Peer-reviewed |
| DSIPdelta sleep-inducing peptide | No reliable published human pharmacokinetic data | No human plasma study published | The frequently quoted 15 minutes is the rate of proteolytic tryptophan cleavage by brain slices and homogenates in vitro. It is not a human plasma half-life and should not be used as one. | Schoenenberger, Eur Neurol 1984 (PMID 6548966)In vitro only |
| GHRP-2 | 0.55 ± 0.14 hours (about 33 minutes) | Intravenous bolus | Measured in 10 prepubertal children with short stature, not in adults. Worth stating rather than presenting as a generic adult value. | Pihoker et al., J Clin Endocrinol Metab 1998 (PMID 9543135)Peer-reviewed |
| GHRP-6 | 2.5 ± 1.1 hours (elimination); 7.6 ± 1.9 minutes (distribution) | Intravenous bolus | Nine healthy males, averaged across 100, 200 and 400 mcg/kg dose levels. The two-phase profile is why a single number understates the early drop. | Cabrales et al., Eur J Pharm Sci 2013 (PMID 23099431)Peer-reviewed |
| GlutathioneGSH · reduced glutathione | 14.1 ± 9.2 minutes | Intravenous infusion | Ten healthy volunteers given 2 g/m² intravenously. The very wide standard deviation is part of the finding. | Aebi et al., Eur J Clin Invest 1991 (PMID 1907548)Peer-reviewed |
| GonadorelinGnRH · Factrel | 2 to 4 minutes | Intravenous | This figure describes endogenous GnRH. Gonadorelin is the identical synthetic decapeptide, so it transfers, but it is not a measurement of the product. The human Factrel product is discontinued and the surviving DailyMed entry is a veterinary label with no pharmacokinetics section. | Lee & Kelly, F&S Reports 2023 (PMC10201293)Review article |
| HCGhuman chorionic gonadotropin · Pregnyl | 29 ± 6 hours terminal (subcutaneous); 26.5 ± 2.5 hours (intravenous) | Subcutaneous or intravenous | These values are for recombinant choriogonadotropin alfa. The serum profile is two-compartment with an initial half-life of 4.5 ± 0.5 hours. Urinary hCG products such as Pregnyl carry no pharmacokinetics section on their labels at all. | Ovidrel label, DailyMedFDA label |
| Hexarelin | No reliable published human pharmacokinetic data | Not established in any species | The widely quoted 55 minutes is not hexarelin's half-life. In the source study it is the decay of the growth hormone response to hexarelin, which is a different quantity. That mix-up has been copied into secondary sources including recent reviews. No pharmacokinetic study of the peptide itself was found in any species. | Imbimbo et al., Eur J Clin Pharmacol 1994 (PMID 7957536)No human PK |
| HGH Fragment 176-191hGH frag 176-191 · GH lipolytic fragment | No reliable published human pharmacokinetic data | Not published | The same molecule family as AOD-9604 and the same gap. Human trials were conducted without a published pharmacokinetic half-life. | Wilding, Curr Opin Investig Drugs 2004 (PMID 15134286)No human PK |
| Human Growth HormoneHGH · somatropin | 7 to 10 hours apparent (subcutaneous); 21.1 minutes (intravenous) | Subcutaneous (apparent); intravenous infusion | The label states explicitly that the longer subcutaneous figure is due to slow absorption from the injection site, so it is absorption-rate-limited rather than a true elimination half-life. The 21.1 minute intravenous value is the closest label-sourced proxy for growth hormone itself. | Norditropin label, DailyMedFDA label |
| Humanin | No reliable published human pharmacokinetic data | No human study published | Rodent intraperitoneal only, and the paper reports a comparison between rats and mice rather than a transferable numeric value. Its title is explicit that the work is in male rodents. | Chin et al., Endocrinology 2013 (PMID 23836030)Animal data only |
| IGF-1 LR3 | No reliable published human pharmacokinetic data | No human study published | No human study of IGF-1 LR3 was found. For contrast, the label for mecasermin, which is recombinant native IGF-1, gives a mean terminal half-life of 5.8 hours after subcutaneous administration in children. LR3 is modified specifically to reduce binding-protein affinity, so the native figure should not be read across. | Increlex (mecasermin) label, DailyMedNo human PK |
| Ipamorelin | Approximately 2 hours (terminal) | Intravenous, 15-minute infusion | Healthy males across five dose levels, dose-proportional, clearance 0.078 L/h/kg. | Gobburu et al., Pharm Res 1999 (PMID 10496658)Peer-reviewed |
| Kisspeptin | 27.6 ± 1.1 minutes (kisspeptin-54) | Intravenous infusion | This value is for kisspeptin-54. No human half-life has been published for kisspeptin-10; the 4 minute figure often quoted for it was measured in male mice. | Dhillo et al., J Clin Endocrinol Metab 2005 (PMID 16174713)Peer-reviewed |
| LiraglutideVictoza · Saxenda | Approximately 13 hours | Subcutaneous | Saxenda's label gives the same 13 hours. This is what supports once-daily rather than once-weekly administration. | Victoza label, DailyMedFDA label |
| LL-37 | No reliable published human pharmacokinetic data | Not established systemically | A human phase 2b trial exists, but it is topical treatment of venous leg ulcers with no systemic pharmacokinetic measurement. | Mahlapuu et al., Wound Repair Regen 2021 (PMID 34687253)No human PK |
| MazdutideIBI362 · LY3305677 | 150.9 to 403.5 hours (6.3 to 16.8 days) | Subcutaneous | A dose-cohort range from non-compartmental analysis after a first dose, not a single mean. A later higher-dose trial reports a considerably wider spread, up to 1075.7 hours, which reflects long terminal tails rather than a different molecule. | Ji et al., eClinicalMedicine 2021 (PMID 34430840)Peer-reviewed |
| Melanotan II | No reliable published human pharmacokinetic data | No human study published | Rat intravenous pharmacokinetics showing biphasic disposition. The human pilot phase 1 study reported tanning, erections and toxicity, but no pharmacokinetic half-life. | Ugwu et al., Biopharm Drug Dispos 1994 (PMID 7981427)Animal data only |
| Mod-GRF (1-29)CJC-1295 no-DAC | No reliable published human pharmacokinetic data | Not established | No controlled human pharmacokinetic study of the tetra-substituted non-DAC analogue exists. The sermorelin value covers unmodified GHRH(1-29) only, and the CJC-1295 value covers the DAC form only. | Teichman et al., J Clin Endocrinol Metab 2006 (PMID 16352683)No human PK |
| OxytocinPitocin | About 1 to 6 minutes | Intravenous or intramuscular | The label notes the half-life decreases further in late pregnancy and during lactation. A generic oxytocin label states 3 to 5 minutes, so the two approved labels genuinely disagree. | Pitocin label, DailyMedFDA label |
| Pentosan PolysulfateElmiron | 27 hours (300 mg) and 20 hours (450 mg) | Oral | Read this one carefully. It is the decline of total plasma radioactivity after a radiolabelled oral dose, not of intact drug, and only about 6% of an oral dose is absorbed. It is not a clean elimination half-life. | Elmiron label, DailyMedFDA label |
| PT-141bremelanotide · Vyleesi | Approximately 2.7 hours (range 1.9 to 4.0) | Subcutaneous | Mean terminal half-life after a single subcutaneous dose of bremelanotide. | Vyleesi label, DailyMedFDA label |
| Retatrutidetriple agonist | Approximately 6 days | Subcutaneous | Phase 1b dose-escalation study of LY3437943. Pharmacokinetics were dose proportional. | Coskun et al., Lancet 2022 (PMID 36354040)Peer-reviewed |
| Selank | No reliable published human pharmacokinetic data | No human study published | Rat tissue distribution using tritium labelling. No human pharmacokinetic study was retrievable. | Ashmarin et al., Bioorg Khim 2008 (PMID 18695718)Animal data only |
| SemaglutideOzempic · Wegovy | Approximately 1 week | Subcutaneous | The label adds that semaglutide remains in circulation for about 5 weeks after the last dose. An independent phase 1b trial measured 145 to 165 hours, which agrees with the label. | Ozempic label, DailyMedFDA label |
| Semax | No reliable published human pharmacokinetic data | No human study published | Rat intranasal distribution study using tritium labelling. Despite Semax being a registered medicine in Russia, no human plasma pharmacokinetic study was retrievable. | Shevchenko et al., Bioorg Khim 2006 (PMID 16523722)Animal data only |
| Sermorelin | 4.3 ± 1.4 minutes | Intravenous infusion | A disappearance half-time measured during constant intravenous infusion in 10 normal men, not a classic post-dose elimination half-life. The widely repeated 11 to 12 minute figure attributed to the Geref label could not be verified; Geref was withdrawn in 2002 and its label is not retrievable on DailyMed or accessdata. | Soule et al., J Clin Endocrinol Metab 1994 (PMID 7962295)Peer-reviewed |
| SS-31elamipretide | No reliable published human pharmacokinetic data | Not stated on the label | Elamipretide is now FDA-approved as Forzinity. Its clinical pharmacology section covers absorption, distribution, metabolism and excretion in detail but states no half-life value anywhere. The roughly 4 hour figure sometimes quoted comes from a canine heart-failure model. | Forzinity (elamipretide) label, FDANo human PK |
| SurvodutideBI 456906 | No reliable published human pharmacokinetic data | Not published | Multiple phase 1 and phase 2 trials of BI 456906 have been published. Each describes the C18 diacid as a half-life extending principle supporting once-weekly dosing, and none publishes a numeric human half-life. | Phase 1 multiple rising dose trial (PMID 36974349)No human PK |
| TB-500thymosin β4 fragment | No reliable published human pharmacokinetic data | Not established | Two points. The human intravenous trials of full-length thymosin β4 describe an increasing half-life with increasing dose but publish no number. Separately, TB-500 as sold is the Ac-LKKTETQ fragment rather than full thymosin β4, and no pharmacokinetic study of that fragment was found in any species. | Ruff et al., Ann N Y Acad Sci 2010 (PMID 20536472)No human PK |
| TesamorelinEgrifta | 8 minutes (Egrifta SV), 11 minutes (Egrifta WR) | Subcutaneous, single dose, healthy subjects | An earlier Egrifta label reported 26 minutes in healthy subjects and 38 minutes in HIV-infected patients after 14 consecutive days. The difference is formulation and repeat dosing, not route. Egrifta has never had an intravenous presentation. | Egrifta SV label, DailyMedFDA label |
| Thymosin α-1Zadaxin | Approximately 2 hours (serum) | Subcutaneous | From a review rather than a primary pharmacokinetic trial. Peak serum concentrations within two hours, returning to baseline within 24 hours. Thymosin alpha-1 is not FDA-approved, so no US label exists to check this against. | Billich, Am J Health Syst Pharm 2001 (PMID 11381492)Review article |
| Thymosin β-4 | No reliable published human pharmacokinetic data | Not established | Two human intravenous trials exist, at 42 to 1260 mg and with recombinant material respectively. Neither reports a numeric half-life in the retrievable record; the later trial reports Cmax and AUC only. | Ruff et al., Ann N Y Acad Sci 2010 (PMID 20536472)No human PK |
| TirzepatideMounjaro · Zepbound | Approximately 5 days | Subcutaneous | Zepbound's label states approximately 5 to 6 days in patients with overweight or obesity. Same molecule, slightly different labelled population. | Mounjaro label, DailyMedFDA label |
| VIPvasoactive intestinal peptide | About 1 minute | Intravenous infusion | Plasma levels fell by first-order kinetics with an average disappearance half-time of one minute in four healthy volunteers. | Domschke et al., Gut 1978 (PMID 730072)Peer-reviewed |
What a half-life is, and what it is not.
The elimination half-life is the time it takes for the concentration of a compound in plasma to fall by half. Because decay is exponential rather than linear, the useful consequence is a rule of thumb: after roughly four half-lives about 94% has gone, and after about five roughly 97%. That is why clinicians talk about five half-lives as practical washout, and it is the arithmetic behind the clearance estimator.
What it does not describe is how long an effect lasts. Plasma concentration and biological effect come apart constantly. A compound can be cleared from blood while its receptor remains occupied, or while a downstream cascade it triggered is still running. Growth hormone secretagogues are the obvious case: the peptide is gone in minutes, and the growth hormone pulse it provoked outlasts it several times over. Confusing those two curves is where the hexarelin number in circulation came from.
Route matters just as much as the molecule. Intravenous administration measures elimination directly. Subcutaneous administration measures whatever is slower, and for a depot that is usually absorption, so the apparent half-life reflects how fast the drug leaves the injection site rather than how fast the body destroys it. This is called flip-flop kinetics, and it is why the route column sits next to every value here rather than being tidied away.
Why most of the catalogue is not in this table.
This site profiles 69 compounds and this table has 42 entries. The other 27 are absent for one reason: we could not find a published pharmacokinetic value for them that traced back to a primary source, in humans or in any species. Not a weak study, not a contested one. Nothing we were willing to cite.
Human pharmacokinetics is expensive work. It needs dosed volunteers, serial blood sampling, a validated bioanalytical assay and compartmental modelling, which in practice means it happens inside a drug-development programme and almost nowhere else. Compounds that never entered one never generated the data, and no amount of online confidence changes that.
The compounds we did include without a value are the ones where a specific number is widely repeated. Leaving them out entirely would let the internet keep the last word, so they are listed with the citation showing what the evidence actually is, whether that is a rodent study, a human trial that never reported pharmacokinetics, or a figure that was measuring something else altogether.
FDA prescribing information first, then peer-reviewed pharmacokinetic studies indexed on PubMed. Vendor product pages, telehealth marketing, forum posts and peptide guide articles were disqualified as sources, no matter how precise the number they quoted. Where a value could not be traced back to a primary study, the cell was left empty. An empty cell is a finding, not a gap we ran out of time to fill.
FAQ.
What does a half-life actually tell me?
It is the time for the concentration in plasma to fall by half. That is all it is. It describes how fast the body clears a compound, which shapes how long a measurable amount persists and how quickly repeated administration reaches a steady state. It says nothing about how long an effect lasts, because effect depends on receptor binding, downstream signalling and tissue distribution, not on plasma concentration alone.
Is half-life a dosing schedule?
No, and this is the single most common misreading of this table. Labelled dosing intervals are set by clinical trials that measured outcomes and tolerability, not by dividing or multiplying a half-life. Some drugs are given far more often than their half-life suggests and some far less. Treating a half-life as a timing instruction is exactly the kind of reasoning this page exists to discourage.
Why do so many rows say no reliable published human data?
Because that is the truth for most research peptides. Human pharmacokinetics requires a clinical study with serial blood sampling and a validated assay, which is expensive and essentially only happens inside a drug-development programme. Compounds that never entered one have no human PK, and sometimes no PK in any species. The numbers circulating for them online are supplier copy or extrapolations from rodents.
You list a rodent value for BPC-157. Why not just use it?
Because rodent clearance does not transfer to humans. Small animals clear peptides faster, they differ in plasma peptidase activity, protein binding and renal function, and interspecies scaling for peptides is unreliable even when done formally. The rat figure is cited so you can see what does exist and how far it is from what you would need, not as a substitute for a human number.
Why is hexarelin listed without a value when a number is widely quoted?
Because the widely quoted figure of about 55 minutes is not hexarelin's half-life. It is the decay of the growth hormone response to hexarelin, measured in the original study, which is a different quantity entirely. That mix-up has been copied into secondary sources including recent reviews. We are not going to propagate it, and no pharmacokinetic study of hexarelin itself was found in any species.
How were these values chosen?
FDA prescribing information via DailyMed first, then peer-reviewed pharmacokinetic studies indexed on PubMed. Vendor pages, telehealth marketing, forum posts and peptide guide blogs were disqualified as sources regardless of how confident they sounded. Where two sources disagreed or a value could not be traced to a primary study, the row was left without a value rather than filled in.
Do half-lives differ by route?
Substantially, and the route column matters as much as the number. A subcutaneous depot absorbs slowly, so the apparent half-life reflects absorption rather than elimination, a phenomenon called flip-flop kinetics. Tesamorelin is the clearest example on this page: minutes when given intravenously, considerably longer when given subcutaneously, for the same molecule.