Peptides for sleep: ranked by what the human evidence supports

Most guides rank peptides for sleep by how loudly the compound is marketed. The published evidence runs close to the reverse. DSIP (delta sleep-inducing peptide) carries the most literal name in the category and the strangest file behind it. It was isolated in 1977 from blood draining out of a rabbit's brain, no one has since identified its gene, its protein or its receptor ¹, and yet it is the only compound here that was ever put in a sleep laboratory to see whether it helps people sleep ^12131415. The compounds with modern overnight brain-wave data are the growth hormone secretagogues, drugs that make the pituitary release growth hormone, and in the one trial that recorded the sleep EEG (electroencephalogram, a recording of brain waves through the scalp) that class made sleep lighter rather than deeper ².

The sleep-peptide market sells anxiety compounds as sleep compounds

Selank is the compound most often recommended for sleep in peptide communities, and its human trial record is an anxiety record. The 2008 trial ran 62 patients with generalized anxiety disorder and neurasthenia (an older diagnosis for chronic nervous exhaustion), 30 on Selank and 32 on medazepam (a benzodiazepine), and the outcomes were the Hamilton, Zung and Clinical Global Impression anxiety scales ⁷. The trial recorded no sleep diary, no actigraphy (a wrist sensor that estimates sleep from movement) and no overnight sleep study. It also recorded psychostimulant effects alongside the anxiolytic ones, which is an odd property for something taken at bedtime.

The gap matters because the two problems come apart. Someone who lies awake with racing thoughts has an anxiety problem that shows up at night, and the anxiety literature is at least relevant to that pattern, though relevance is not evidence of benefit. Someone who falls asleep in four minutes and wakes unrefreshed after eight hours has a sleep-architecture problem, and nothing in the Selank literature speaks to it. Naming that gap is most of what this page is for.

Ranked by the strength of human sleep evidence

The table below ranks the options by how close the best available study gets to a human sleep outcome, which puts two non-peptide comparators at the top and the compound with the emptiest molecular file third. That ordering is uncomfortable for a site that writes about peptides, and it is what the literature supports.

1

SelankThe anxiolytic peptide2

EpithalonThe telomere peptide3

DSIPThe deep-sleep peptide4

CJC-1295The GH amplifier5

IpamorelinThe selective GH releaser

Browse the full reference library →

Does Selank help you sleep?

No trial has answered that question directly, because no Selank trial has used a sleep measure as its outcome. The closest published finding runs opposite to the way it is usually quoted. In a 2015 Russian anxiety trial, adding Selank to the benzodiazepine phenazepam reduced that drug's unwanted effects, including sedation and the increase in sleep duration it caused, in 40 patients against 30 on phenazepam alone ¹¹. That is Selank cutting benzodiazepine oversleeping, not Selank improving sleep, and it has never been compared against any sleep aid.

Selank is an analog of tuftsin, a peptide the human immune system already makes, and it is given as a nasal spray in community use. The only hypothesis its literature supports is narrow: sleep trouble that sits downstream of bedtime anxiety, racing thoughts, a mind that will not stop rehearsing tomorrow. Whether that hypothesis applies to any individual person, and whether acting on it makes sense at all, is a conversation for a clinician rather than a call this page can make. If sleep onset is fine and the problem is waking at 3 a.m. or waking unrefreshed, the Selank literature has nothing to offer. Full detail on the compound sits on the Selank page, and the anxiety side of the evidence is on peptides for anxiety. Semax, Selank's better-known sibling, is the wrong tool at bedtime, since it is used as a stimulant-type nootropic; the two are set side by side on Semax vs Selank and on the Semax page.

Sleep peptides and DSIP: the compound named after deep sleep, and the trials everyone forgot

DSIP was named for the delta waves it was supposed to induce, and the name has been doing the persuading ever since. The peptide was isolated in Basel in 1977 from blood draining out of the brain of a rabbit whose thalamus had been electrically stimulated, and the sequence turned out to be a nonapeptide, nine amino acids long. As of the most recent review of the literature, in 2006, no DSIP gene, protein or receptor had been isolated ¹, and no later report of one has surfaced.

The hypothesis regarding DSIP as a sleep factor is extremely poorly documented and still weak ¹.Vladimir Kovalzon and Tatyana Strekalova, Journal of Neurochemistry, 2006

The same review flags a detail that most DSIP sales copy skips: slow-wave-sleep-promoting activity in rabbits and rats appeared with certain artificial structural analogs of DSIP rather than with DSIP itself ¹. The compound sold under the name is not the compound that produced the animal result.

The human trials are the part both the sellers and the skeptics get wrong. Between 1981 and 1992, synthetic DSIP was infused intravenously at 25 nmol/kg into small groups in sleep laboratories with polysomnography running. The first reports were positive: 6 middle-aged chronic insomniacs slept longer with fewer interruptions and no daytime sedation ¹², and 6 healthy volunteers given a morning infusion showed shorter sleep onset and better sleep efficiency the following night ¹³. The two double-blind trials that followed did not hold the line. A 1987 crossover trial in chronic insomniacs found total sleep time and stage 2 sleep increased while slow-wave sleep did not move, and concluded that the improvement was of little clinical significance ¹⁴. A 1992 double-blind parallel-group trial in 16 chronic insomniacs found higher sleep efficiency and shorter sleep latency, judged the effects weak and possibly driven by a change in the placebo group, and concluded that short-term DSIP treatment is unlikely to be of major therapeutic benefit ¹⁵. No human DSIP sleep trial has been published since.

So DSIP is neither the empty compound its molecular file suggests nor the breakthrough its name suggests. It is a 1980s intravenous research drug that produced small, inconsistent effects in fewer than 50 people in total and has been left alone for 34 years, while the vials sold today are for injection under the skin at doses nobody published. The DSIP page carries the rest of the pharmacology.

Do growth hormone peptides deepen sleep, or lighten it?

Growth hormone release and deep sleep are genuinely linked in human physiology, which is why ipamorelin, CJC-1295 and their relatives get sold with sleep claims attached. When researchers put that link to the test with an EEG recorder running, the result went the other way. Frieboes and colleagues gave 7 healthy young volunteers 4 doses of 50 mcg hexarelin, a synthetic growth hormone secretagogue in the same class as ipamorelin, across a night in the sleep laboratory. Stage 4 sleep fell in the first half of the night, EEG delta power fell across the whole night, and growth hormone, prolactin, ACTH and cortisol all rose ².

Sleep is not the only question this class leaves open, and the rest of it belongs on the same page. Raising growth hormone and IGF-I has a documented side-effect profile in people: a systematic review of growth hormone given to healthy elderly adults found significantly more soft tissue swelling, joint pain, carpal tunnel syndrome and gynecomastia, and somewhat more impaired fasting glucose and new-onset diabetes ¹⁶. Sustained IGF-I elevation is also why active or suspected cancer excludes people from growth hormone treatment. Those findings come from growth hormone therapy rather than from measurements taken on ipamorelin, which nobody has studied that way, and a compound whose entire purpose is to raise growth hormone inherits the question.

The same Munich group also found that sex flipped the direction of the effect. Across 35 patients with depression and 40 matched controls, GHRH given as 4 doses of 50 mcg overnight enhanced non-REM sleep and stage 2 sleep in men regardless of diagnosis and decreased both in women, a difference large enough to be statistically significant ⁴. Any protocol that promises the same sleep outcome to everyone is ignoring one of the few sleep-EEG datasets that exists for this class.

CJC-1295 belongs in this section for what it does not show, and for a labeling problem. Teichman and colleagues studied CJC-1295 with DAC (drug affinity complex), the version engineered to bind albumin, which is where the estimated half-life of 5.8 to 8.1 days comes from; single ascending doses and repeated weekly doses in healthy adults raised growth hormone and IGF-I (insulin-like growth factor 1, the hormone through which most of growth hormone's effects are delivered) and measured nothing about sleep ⁸. Most grey-market product sold as CJC-1295 is the no-DAC form, modified GRF 1-29, whose half-life is measured in minutes, so the published pharmacology does not transfer to what people are buying. Details sit on the CJC-1295 page and the ipamorelin page, and the genetics angle on growth hormone peptides and sleep is covered in ipamorelin, sleep and genetics.

Is epitalon good for insomnia?

Epitalon is a four-amino-acid peptide claimed to restore pineal gland function, and the pineal gland is what makes melatonin. The load-bearing study is Goncharova and colleagues in 38 female rhesus macaques, 18 of them young at 6 to 8 years and 20 aged at 20 to 26 years. Epitalon at 10 micrograms per animal per day, intramuscularly for 7 to 10 days, raised night melatonin in the aged animals and did nothing in the young ones ⁹. The nearest human-adjacent data is a melatonin-rhythm measurement in old monkeys and elderly people, and it covers epithalamin, a bovine pineal peptide complex, alongside epitalon, the synthetic tetrapeptide ¹⁰. Both are hormone readings rather than sleep outcomes. No published trial has measured what epitalon does to sleep, and none has reported adverse-event monitoring in people as an endpoint.

Where epitalon is plausible, if it is plausible anywhere, is the phase-advanced pattern that comes with age: asleep at 9 p.m., wide awake at 4 a.m., melatonin rhythm flattened. It is a poor fit for acute insomnia, and whatever it does would take cycles rather than a night. The epithalon page covers the longevity claims separately. Morning light exposure at a consistent time has evidence for circadian shifting in its own right and costs nothing.

Do collagen peptides help sleep?

Collagen peptides are the highest-volume search in this category and the weakest link to sleep. No human trial has tested collagen peptides with a sleep outcome. The one mechanism linking collagen to sleep-relevant biology runs the other way.

Tryptophan is the amino acid the body converts into serotonin and then into melatonin, so a large collagen load taken in place of a complete protein is mechanistically the same maneuver researchers use when they want to lower serotonin precursor availability. The full mechanism, the animal work behind it and the 5-HTTLPR genotype nuance are on peptides for anxiety. Nothing here says collagen wrecks sleep; it says the answer to whether collagen peptides help sleep is that the only relevant mechanism points the other way.

Why the same sleep peptide works for one person and not another

Two people can run an identical protocol and report opposite results, and part of the reason is set before any compound is involved. Rétey and colleagues showed that a functional variant of adenosine deaminase (ADA, rs73598374), the enzyme that breaks adenosine down to inosine, increases both the duration and the intensity of deep sleep in humans ⁵. Adenosine is the molecule that accumulates while you are awake and creates sleep pressure; slower breakdown means more of it, and measurably deeper sleep.

The same PNAS paper found that a polymorphism of the adenosine A2A receptor gene (ADORA2A) was associated with differences between people in anxiety symptoms after caffeine, and affected the EEG in both sleep and wakefulness ⁵. One pathway therefore connects how deeply you sleep and how anxious caffeine makes you feel. For a lot of people with mild insomnia, the highest-yield experiment is an earlier caffeine cutoff rather than a vial of anything.

How people dose these compounds, and what the studies did

Two different things get conflated in every protocol thread: what researchers administered in a published study, and what people report doing at home. They are separated below on purpose. Neither is an instruction, and the published protocols were designed to answer a research question rather than to treat anyone's insomnia.

What community protocols report, and why they are not rules

• Read this list as reporting, not as a protocol. These figures describe what forum users say they do, reproduced so the gap between community practice and the research record is visible. No clinical study established any of these doses, routes or schedules, none was monitored for safety, the material is unapproved and its purity is unverified, and nothing on this page is dosing advice.
• Selank: 300 to 900 mcg intranasally, on a 5-nights-on and 2-off cycle. The cycling convention comes from forums rather than from a trial; no Selank study has measured tolerance, so the absence of tolerance reports is not evidence that tolerance does not occur.
• Epitalon: 5 to 10 mg subcutaneously, roughly a thousand times the per-animal epitalon dose used in the monkey study, on a 10-day schedule borrowed from that same study. No dose-finding or safety study in people supports either the amount or the route.
• DSIP: deliberately absent. Every published human dose was 25 nmol/kg given intravenously in a hospital sleep laboratory, which is not what anyone is doing at home, so the subcutaneous numbers circulating in forums have no published basis at any dose.
• Growth hormone peptides: usually injected at night, on the theory that the growth hormone pulse tracks deep sleep. The one trial that measured what happens to the EEG in this class found less deep sleep rather than more ².
• Timing complaint that comes up repeatedly: intranasal Selank close to lights-out can feel alerting, which fits the psychostimulant effects recorded in the 2008 trial ⁷.

Cost, storage, and what is in the vial

Grey-market listings for these compounds run $25 to $60 for a single vial of 5 to 10 mg, with nasal-spray Selank at the upper end. Treat those figures as an observation of what vendors advertise rather than a cited fact, because nothing here is priced through a regulated supply chain.

• Purity is the real cost variable. No regulator verifies grey-market peptide purity and no independent assay program covers these compounds, so a certificate of analysis supplied by the seller is the only document in existence and it is supplied by the seller.
• Lyophilized (freeze-dried) vials are generally stored refrigerated and kept out of light; once mixed with bacteriostatic water, refrigeration and a limited working window apply. No independent stability data exists for these specific compounds, so shelf-life claims on vendor pages are unverified.
• Reconstitution is where most beginners make errors that change the dose by a factor of ten. The mechanics are covered in how to reconstitute peptides, and sterile technique in the peptide injection hygiene guide. Those guides exist as harm-reduction education for people who have already made a decision with a clinician; they are not an endorsement of using any compound on this page.
• If you want to see how vendors compare on documentation and third-party testing, our vendor comparison lays out what each one publishes. PeptidesDNA earns a commission if you buy through links on that page. The commission does not change what the research says, and none of these compounds has FDA approval for human use.
• Time to any measurable answer is longer than most people budget. The epitalon monkey work ran 7 to 10-day cycles and looked at melatonin over months; the Selank anxiety trial measured anxiety scales over weeks. A three-night experiment answers nothing.

Are peptides for sleep legal in the US?

None of these compounds has FDA approval for any human use, and the restrictions in play govern what licensed compounding pharmacies may prepare rather than what an individual may possess. Selling is a different question. Because none of them is an approved drug, marketing them for human use violates the Federal Food, Drug, and Cosmetic Act, which is why grey-market vendors label vials "for research use only" and why FDA has detained peptide shipments at import. The research-use label does not make human use lawful, and it does not make the contents verified or safe.

The compounding picture has moved twice since 2023 and most peptide pages still quote the old version. FDA placed several of these substances in Category 2 of its interim 503A bulk drug substances list in September 2023, citing concerns including immunogenicity risk (the chance the body mounts an immune reaction to the injected peptide) and unnatural amino acids that make a substance hard to characterize. Ipamorelin acetate, CJC-1295 and selank acetate left Category 2 in September 2024 when the nominators withdrew their nominations, and epitalon and emideltide (DSIP) were removed in April 2026 ahead of a committee review. A withdrawn nomination is not a safety clearance, and leaving Category 2 does not authorize anything.

What happened next is the more informative part. FDA's Pharmacy Compounding Advisory Committee voted against adding ipamorelin acetate to the 503A bulks list on 29 October 2024, and against CJC-1295 in all its forms on 4 December 2024. At the 23 to 24 July 2026 meeting, the same committee reviewed emideltide (DSIP) for insomnia, narcotic dependence and opioid withdrawal, and epitalon for insomnia. It recommended epitalon along with five other peptides, and rejected emideltide, the only rejection of the two days. Those votes are non-binding in both directions, and adding any substance to the 503A list still requires a formal public rulemaking process that has not started.

Peptides for sleep apnea: what to rule out first

Sleep apnea is an airway problem, and no peptide in this article addresses airways. Someone with untreated apnea who takes a compound for unrefreshing sleep is treating the symptom of a mechanical obstruction that keeps interrupting their breathing all night. The diagnostic is a sleep study, and the treatments that work are structural: CPAP, a dental device, positional changes, weight change, sometimes surgery.

Further reading inside PeptidesDNA: the ranked roster with the full evidence grading sits in best peptides for sleep, the anxiety half of this cluster is on peptides for anxiety, and anyone new to the category should start with peptides for beginners.