Best peptides for sleep (2026): five candidates, ranked honestly by what the research actually shows.
Ranking peptides by popularity would put DSIP first — it is the "sleep peptide" by name and has dominated the research-chemical sleep category for decades. Ranking by what the published evidence actually supports produces a different order. This comparison ranks five candidates by evidence strength: human polysomnography data, mechanistic clarity, side-effect profile, and the quality of the evidence-gathering process. The ranking is honest even when the result is counterintuitive.
- MK-677: Strongest evidence (Copinschi 1997 PSG, GHRP-6 class confirmation). Indirect mechanism via GH-pulse amplification. Meaningful side-effect profile.
- CJC-1295 / Ipamorelin: Same GH/SWS mechanism as MK-677, no direct PSG trial. Extrapolated evidence, not confirmed.
- Epitalon: Human melatonin data from Khavinson group. Plausible circadian mechanism. No independent Western replication.
- Selank: Documented anxiolytic. Indirect sleep benefit in anxiety-driven insomnia. Not a primary sleep compound.
- DSIP: Category namesake. No primary receptor identified. Human data small and unreplicated. The most popular choice in this category has the weakest mechanistic foundation.
How this ranking was made
A compound rises in this ranking to the extent that it satisfies four criteria:
- Human polysomnography (PSG) data — objective, pre-specified, measurement of sleep architecture in human subjects. PSG is the gold standard for sleep research; subjective sleep-quality ratings are supporting evidence, not primary data.
- Identified primary mechanism — a confirmed receptor or molecular target that explains how the compound produces its sleep effect. Without a primary mechanism, dose-response relationships cannot be defined.
- Independent replication — findings confirmed by research groups other than the compound's developers, in different countries, with pre-specified protocols.
- Proportionate side-effect profile — the off-target effects must be proportionate to the sleep benefit, especially for a compound being used specifically for sleep rather than for a primary indication.
This ranking is not a clinical recommendation, a protocol guide, or an endorsement of any compound for human use. All compounds discussed are research chemicals sold under a research-use-only framing. The full Sleep & Circadian pillar page contains the evidence hierarchy across all twenty-eight spokes.
Rank 1 — MK-677 (Ibutamoren)
Evidence grade: Phase II human PSG (secondary endpoint). MK-677 is a ghrelin receptor agonist (GHSR-1a) that stimulates pulsatile GH release from the pituitary. The mechanism links directly to sleep: nocturnal GH secretion is tightly coupled to slow-wave sleep (SWS), and amplifying the GH pulse amplifies SWS. This is not a theoretical extrapolation — it is confirmed by decades of sleep endocrinology research (Steiger 2002).
The sleep-specific evidence: Copinschi et al. (Neuroendocrinology, 1997;66(4):278–286) — 8-week randomised placebo-controlled crossover with full PSG — showed increased stage IV and REM sleep, unchanged sleep onset latency. Frieboes et al. (Arch Gen Psychiatry, 1995;52(11):958–961) showed the same SWS enhancement with IV GHRP-6 (same receptor class). Two studies, two designs, one mechanism — the class-level evidence is as strong as anything in this category.
Limitations: MK-677 increases appetite substantially (ghrelin is the hunger hormone), causes water retention, and may impair insulin sensitivity with chronic use. These are not trivial. A compound that improves sleep while also increasing caloric intake and fluid retention is not a clean sleep intervention. Sleep was a secondary endpoint in Copinschi — no pre-registered Phase III sleep trial exists.
Full mechanistic analysis: MK-677 sleep architecture spoke. Broader pharmacology: MK-677 overview in Pillar 3.
Rank 2 — CJC-1295 / Ipamorelin (GH-axis stack)
Evidence grade: mechanistic extrapolation — no direct PSG trial. CJC-1295 is a GHRH analog; Ipamorelin is a selective ghrelin receptor peptide. Both amplify GH pulsatility through the same GH/SWS coupling mechanism that gives MK-677 its sleep effect. The rationale for SWS enhancement is mechanistically identical to MK-677; the difference is that no dedicated human sleep study (PSG as primary or secondary endpoint) has been published for either compound.
The evidence is extrapolated from: (1) the confirmed GH/SWS coupling in endocrinology literature, and (2) the Copinschi and Frieboes data showing PSG-confirmed SWS enhancement with other GHSR agonists. This extrapolation is reasonable but it remains extrapolation. CJC-1295 and Ipamorelin rank above Epitalon and DSIP because their mechanism is better defined — they act on known, cloned receptors with quantified pharmacology — but below MK-677 because the sleep-specific human data does not exist for them.
Ipamorelin in particular has a cleaner side-effect profile than MK-677: it is selective for GH release with minimal appetite stimulation, minimal cortisol/prolactin elevation, and less water retention. If the sleep benefit of GH-pulse amplification is real and class-wide, Ipamorelin's better tolerability profile may make it preferable to MK-677 for researchers focused specifically on sleep — but this inference rests on unconfirmed premises.
Related spokes: peptides for deep sleep (6.9), CJC-1295 and sleep quality (6.15), Ipamorelin sleep spoke (6.20). For the underlying physiology: growth hormone and slow-wave sleep (6.14). Cross-pillar: the primary Ipamorelin page at Ipamorelin research page in Pillar 3.
Rank 3 — Epitalon
Evidence grade: human melatonin data, single-group, no independent replication. Epitalon (Ala-Glu-Asp-Gly) targets the pineal gland specifically, proposing to restore melatonin synthesis in aged glands. If the mechanism is correct, it addresses the circadian root cause of age-related sleep disruption rather than amplifying GH pulsatility. This is a different — and in some ways more precise — sleep-relevant mechanism than MK-677's.
The published evidence (Korkushko et al., Neuroendocrinol Lett, 2003; Khavinson et al., 2001) shows melatonin profile normalisation in elderly subjects after Epitalon administration. These studies are real and published in indexed journals. The limitation: they originate from the Khavinson group at the St. Petersburg Institute of Bioregulation — the same group that developed Epitalon. No independent Western research group has replicated these findings in a pre-registered trial.
Epitalon ranks third rather than first on the strength of a plausible mechanism + some human data, despite its lack of a confirmed primary receptor at the molecular level (the DNA-binding mechanism is proposed, not crystallographically confirmed). It ranks above DSIP because the circadian mechanism is at least coherently proposed and the human data show a biologically meaningful endpoint (melatonin level, an objective biomarker). It ranks below MK-677 because the evidence is from a single group without independent replication.
Circadian-specific analysis: Epitalon and melatonin rhythm spoke. Full longevity and anti-aging profile: Epitalon overview in Pillar 3.
Rank 4 — Selank
Evidence grade: human anxiolytic data; sleep benefit is indirect and conditional. Selank (Thr-Lys-Pro-Arg-Pro-Gly-Pro) is a synthetic analog of the endogenous tuftsin peptide. Russian clinical literature documents anxiolytic effects through GABA-A receptor modulation and BDNF pathway upregulation. The sleep relevance is indirect: anxiety is one of the leading causes of sleep-onset insomnia, and an anxiolytic that reduces arousal without sedation could improve sleep quality in subjects where anxiety is the limiting factor.
Selank's position here reflects the conditional nature of its sleep benefit. It is not a sleep peptide in the mechanistic sense — it does not target sleep-specific receptors, does not alter GH pulsatility, and does not act on the pineal gland. Its sleep benefit depends entirely on anxiety being the root cause of sleep disruption in the specific research subject. In subjects without anxiety-driven insomnia, there is no principled reason to expect a sleep benefit. This conditionality limits its rank despite Selank having more robustly characterised anxiolytic pharmacology than DSIP has for any effect.
Selank for sleep: Selank for sleep spoke (6.5). Sleep-anxiety overlap: peptides for sleep anxiety spoke (6.19). Primary Selank pharmacology in Pillar 2: Selank research page.
Rank 5 — DSIP (Delta Sleep-Inducing Peptide)
Evidence grade: pre-Phase II; no primary receptor; small, old, unreplicated human studies. DSIP is the most over-specified compound in this category relative to its actual evidence base. Its name implies a specific, direct, potent sleep-inducing mechanism. The science does not support this framing. As of April 2026:
- No primary receptor for DSIP has been identified. Proposed interactions include opioid receptors, benzodiazepine binding sites, and calcium channels — none confirmed as primary mediator.
- The best human data (Schneider-Helmert and Schoenenberger, 1983) used small samples, older methods, and has not been replicated in adequately powered modern trials.
- No Phase II or Phase III trial for DSIP is on record in the US or EU as of April 2026.
- The bulk of modern DSIP literature is from Russian/Eastern European groups, raising the same geographic-concentration caveat as the Epitalon evidence.
DSIP ranks last not because it definitely does not work — the 1977 Schoenenberger/Monnier discovery (PNAS, 74(3):1282–1286) is genuine science that has never been retracted or fundamentally contradicted. It ranks last because the evidence base required to place it above the other candidates — receptor confirmation, dose-response characterisation, independent PSG replication — does not exist. Popularity in the research-chemical market is not a substitute for evidence.
Full DSIP analysis: DSIP deep-dive spoke (6.1). What remains unresolved: DSIP research gaps spoke (6.28). DSIP vs melatonin: melatonin vs DSIP spoke (6.11).
Comparison table
| Compound | Rank | Mechanism clarity | Best human evidence | Independent replication | Sleep side-effect concern |
|---|---|---|---|---|---|
| MK-677 | 1 | Confirmed GHSR-1a → GH → SWS | PSG RCT (Copinschi 1997) | Yes (Frieboes class confirmation) | High (appetite, fluid, glucose) |
| CJC-1295 / Ipamorelin | 2 | Confirmed receptors → GH → SWS (extrapolated) | No direct sleep PSG trial | Partial (GHRP-6 class) | Lower than MK-677 |
| Epitalon | 3 | Proposed pineal → melatonin (not crystallographically confirmed) | Melatonin normalisation data (Korkushko 2003) | No | Low (data-limited) |
| Selank | 4 | Confirmed GABA-A → anxiolysis → indirect sleep | Russian anxiolytic RCTs | Limited | Conditional on anxiety phenotype |
| DSIP | 5 | No confirmed primary receptor | Small, 1983, unreplicated PSG study | No | Unknown (mechanism unknown) |
What the ranking means for research design
This ranking does not mean MK-677 is the "best" compound in an unqualified sense — it has the most sleep-specific human evidence while also having the most significant side-effect profile. A researcher designing a sleep study would face a genuine trade-off: the compound with the best evidence also comes with appetite stimulation, fluid retention, and potential insulin resistance as confounders and tolerability concerns.
The ranking does mean that DSIP should not be chosen as a sleep-research candidate primarily because of its name or its market prevalence. The category namesake happens to be the least mechanistically characterised compound in the group. A researcher who wants to study something in this cluster with a principled basis should understand the GH/SWS coupling mechanism (see the GH-SWS mechanism spoke) before choosing a compound, and should prioritise mechanism-anchored candidates over historically popular ones.
For researchers specifically investigating age-related sleep disruption with a circadian mechanism rationale, Epitalon's pineal angle is the most targeted — but requires accepting the limitations of single-group Russian evidence. For researchers investigating anxiety-driven insomnia, Selank's anxiolytic-first mechanism may be appropriate, but should not be called a sleep peptide in the same breath as a direct SWS enhancer.