Peptides in skincare: the honest guide to what they do

Peptides have overtaken retinol and niacinamide in US search interest this year, and searches for copper peptides, GHK-Cu and peptide serums keep climbing. Every shelf now has a peptide moisturizer, a multi-peptide serum and a peptide eye cream.

The idea is simple: peptides are short chains of amino acids, the building blocks of proteins like collagen. Some are designed to act as messages to skin cells. The question is how well those messages get through a cream, and how much difference they make.

Most cosmetic peptides fall into a few loose groups. The group tells you what the peptide is supposed to do, not how well it does it.

• Signal peptides: fragments meant to nudge skin into making more collagen. Palmitoyl pentapeptide-4 (Matrixyl) and palmitoyl tripeptide-1 with palmitoyl tetrapeptide-7 (Matrixyl 3000) are the famous ones.
• Carrier peptides: they ferry a trace mineral into skin. Copper tripeptide-1, sold as copper peptides or GHK-Cu, is the main example.
• Neurotransmitter-inhibiting peptides: designed to soften the muscle signals behind expression lines. Acetyl hexapeptide-8 (Argireline) is the best known, which is where the "Botox in a bottle" label comes from.
• Enzyme-inhibiting and other peptides: a mixed group, often from soy or rice proteins, with even less human data.

The best single-peptide trial is for palmitoyl pentapeptide-4: a 12-week, placebo-controlled, split-face study of 93 women found fewer visible fine lines on the peptide side. It was run by the company that developed the ingredient.

Argireline has a randomized, placebo-controlled study of 60 people showing smoother lines around the eyes after four weeks. Copper peptides have strong lab and wound-healing data, but few modern controlled trials of the cosmetic serums people buy.

Reviews of cosmetic peptides keep landing in the same place: promising, generally well tolerated, mostly small and short studies, and much of the work is funded by ingredient makers. Whether enough peptide gets through the skin barrier is still an open question for many of them.

Daily sunscreen has the strongest evidence for preventing visible aging. Retinoids have decades of controlled trials for fine lines. Peptides sit a step below: a gentle add-on rather than the main event.

That gentleness is the real selling point. Peptides rarely sting, they suit sensitive skin, and they layer easily with retinol, niacinamide, hyaluronic acid and vitamin C.

• Pick one peptide product that you'll use every day, usually a serum or moisturizer.
• Keep sunscreen in the morning; it does more for lines than any peptide.
• If you can tolerate it, a retinoid at night will usually do more than adding a second peptide product.
• Give it 8 to 12 weeks before judging, and take photos in the same light.

Collagen is broken down into small fragments as skin repairs itself. Some of those fragments, called matrikines, appear to tell fibroblasts to make more collagen. Signal peptides copy that idea. Lab studies support the mechanism; the size of the effect on a real face, through a cream, is the uncertain part.

Many cosmetic peptides are attached to a fatty acid (the "palmitoyl" part) to help them pass the skin's outer layer.

• A modest improvement in the look of fine lines with a signal peptide such as palmitoyl pentapeptide-4.limited evidence
• Smoother expression lines around the eyes with Argireline over a few weeks.limited evidence
• Daily sunscreen for preventing visible aging.well-supported
• Retinoids for fine lines.well-supported

• “Peptides rebuild collagen like a filler.”

  Topical peptides may support collagen signals modestly. They don't replace lost volume.

  insufficient evidence

• “More peptides in one bottle means better results.”

  There's no good evidence that a 10-peptide blend beats one well-studied peptide at a sensible level.

  insufficient evidence

• Peptides working as well as retinoids on their own.
• Most named peptide blends; many have only in-house manufacturer data.
• Injected peptides bought online, which are a different and riskier thing.