Oligopeptide-68 is the peptide brightening serums use when they want a pigment ingredient that is not vitamin C, niacinamide or an acid. Papers and suppliers describe it as a "TGF-β1 biomimetic", a mimic of a growth factor that slows pigment production in laboratory work. That label can be checked, because the molecule's full structure is public. This page decodes it, then sets out what the two human studies that name it can and cannot support, as of 26 October 2026.
Twelve amino acids, read from the structure
PubChem's record (CID 163285842, CAS 1206525-47-4) gives formula C62H91N17O21 and 1,410.5 daltons, but no sequence line. Reading the residues out of its IUPAC name, from the free amino end, gives:
Gly–Arg–Gly–Asp–Tyr–Ile–Trp–Ser–Leu–Asp–Thr–Gln (GRGDYIWSLDTQ)
We checked that reading by adding up the atoms of those twelve amino acids, less the water lost at each of the eleven bonds: the total is C62H91N17O21, exactly PubChem's formula. So the sequence is not a guess.
Where those twelve come from
Two separate pieces are stitched together:
- YIWSLDTQ, the last eight, is copied verbatim from human TGF-β1. In the UniProt record for the TGF-β1 precursor (P01137) it sits at positions 328 to 335; the mature growth factor begins at position 279, so this is residues 50 to 57 of the active protein, in the stretch that runs …CPYIWSLDTQYSK…. That is what "TGF-β1 biomimetic" actually means here: an eight-residue fragment.
- GRGD, the first four, contains RGD, the three-letter motif that integrins, the cell's adhesion receptors, recognise on proteins such as fibronectin. TGF-β1 has an RGD of its own, but at position 244, in the part of the precursor that is cut away, not next to the fragment above. The pairing in oligopeptide-68 is a design choice, presumably to help the peptide stick to cells.
A fragment is not the protein. TGF-β1 is a 112-residue folded dimer; whether eight of its residues on their own act on pigment cells has to be shown for the fragment, and the published record does not show it.
Size: a large peptide with no tail
At 1,410.5 daltons the peptide is nearly three times the roughly 500-dalton size usually quoted as the easy limit for passive passage through intact skin. Many cosmetic peptides answer that problem with a palmitic-acid tag; oligopeptide-68 has none. No paper we found measured how much of it reaches the basal layer, where pigment cells sit.
The two human studies
PubMed returned 2 records for "oligopeptide-68" on 26 October 2026, the same two under the field-tagged form. Neither is tagged as retracted. Both are covered from the pigment side on our peptides for hyperpigmentation page; here is what each can say about the peptide itself.
| Study | Who and how long | What was applied | Can it isolate oligopeptide-68? |
|---|---|---|---|
| Pratchyapurit 2016, J Cosmet Dermatol, Institute of Dermatology, Bangkok | 40 women with facial melasma, randomized, double-blind, 12 weeks; 38 completed | A night serum with diacetyl boldine, and a day cream with diacetyl boldine, oligopeptide-68 and sunscreen, both on the face | No — every face received the peptide inside a combination, with no peptide-free arm |
| Agozzino 2020, J Clin Aesthet Dermatol, Naples and Rome | 20 women with melasma, three months, assessed by confocal microscopy | One "depigmenting agents complex": oligopeptide-68, phytic, glycolic and lactic acids, and the sunscreen filter octinoxate | No — five actives together, no comparison group described |
Both report improvement. In the Thai study 76.3% of completers graded themselves moderately improved and 2.6% markedly improved, and most had a temporary, mild skin reaction. The Italian study reported improvement in all patients without adverse reactions and declares no funding. Exfoliating acids and daily sunscreen are both established parts of melasma care, so neither result can be credited to the peptide.
The hydroquinone comparison, read closely
The Thai paper is often quoted as showing the peptide formulas beat 4% hydroquinone, the long-standing prescription benchmark. Its own title is more careful: either formula is "equal to or better than 4% hydroquinone on normal skin". The methods compared the formulas with 2% and 4% hydroquinone on sun-protected skin of the arms, not on melasma. A lightening agent's effect on normal arm skin is a different measurement from its effect on a facial pigment disorder, so the study cannot rank the products against hydroquinone for melasma. Ingredient databases that repeat "superior to HQ" without that qualifier overstate the paper.
The supplier's numbers
The ingredient database INCIDecoder (now INKEEDecoder, read 26 October 2026) relays a manufacturer test of a 5% formula of the trade ingredient in 23 Asian volunteers for 56 days, in which 87% reported a more even tone and 91% brighter skin. Those are self-reported percentages from a formula, not measurements of the peptide, and the study is not in PubMed.
On labels
COSRX's The 6 Peptide Skin Booster Serum declares oligopeptide-68 11th of 44, among six consecutive peptides, and the Korean peptide serum census lists it with its two records. No label we have read states the amount; under 1%, US labels may list ingredients in any order. Our guide to reading a peptide ingredient list explains what a position can and cannot tell you.
What can fairly be said
Oligopeptide-68 is a twelve-residue peptide built from an integrin-binding RGD motif and an eight-residue fragment of TGF-β1, which makes its "biomimetic" label literally true and functionally unproven. It is large, untagged and tested in people only inside mixtures with established lightening and exfoliating agents. For persistent pigment, including melasma, a dermatologist can set out options with a far larger evidence base; the trial counts behind those options are on our hyperpigmentation page.
Sources
Opened 26 October 2026: PubChem CID 163285842 (formula, weight, IUPAC name, synonyms); UniProt P01137 (human TGF-β1 precursor sequence; fragment positions computed from it); PubMed searches "oligopeptide-68" (2), oligopeptide-68[tiab] (2); abstracts and conflict statements of PMIDs 26833454 and 33133340; INCIDecoder (INKEEDecoder) oligopeptide-68 entry for the manufacturer's test as it relays it. Sequence and formula check: our arithmetic. Corrections go to the contact page.