Does centella asiatica help skin? What the research supports

Can centella asiatica help skin, and how does it stack up against the marketing?

Bottom line: it may modestly help hydration and irritation, and the marketing runs well ahead of the data. The human evidence is real but thin, and most of it is indirect.

A comprehensive dermatology review found that centella's compounds have been studied in acne, burns, atopic dermatitis and wounds, but mostly through cell and animal signaling work. Its authors call for "additional rigorously controlled long-term clinical trials."

The compounds in question are asiaticoside, madecassoside, asiatic acid and madecassic acid. A 2024 review describes them as boosting collagen synthesis, modulating inflammation and offering antioxidant protection. Those are mechanisms, not proof that the serum on your shelf does any of it. More on the ingredient lives on our centella asiatica topic page.

What is centella asiatica good for in skincare?

Soothing and barrier support is the most realistic claim, and the gains are small. Three studies carry most of it: a 4-week cream and hydrogel study, a double-blind trial in batik workers with dry skin, and an oral extract RCT.

The oral trial's authors called the hydration and elasticity changes modest. Safety markers stayed normal, with no serious adverse events. That study can't back any claim about creams.

If you already use a well-formulated moisturizer, the gains in these studies are too small to promise you'd notice a switch.

What do studies say about centella and wound healing versus everyday skincare?

Healing damaged skin has the strongest evidence, but those trials aren't about everyday creams. A 2022 systematic review found just four clinical trials and said more studies were needed before a meta-analysis could confirm effects.

• Diabetic wounds: 170 people, 3 weeks, oral asiaticoside. Wound contraction beat control (day 7: 28.57% vs 12.79%; day 14: 38.10% vs 18.60%), but the placebo wasn't specified.
• Burns: a 3% topical product vs silver sulfadiazine, an active comparator rather than placebo. The review lists 75 participants; another review says 60, so the primary paper needs checking.
• Chronic anal fissure: 98 people, oral plus topical vs no treatment, unblinded and rated high risk for blinding.
• Post-laser skin: 30 people, 0.05% ECa 233 (a standardized extract, roughly 51% madecassoside and 38% asiaticoside) vs placebo for 12 weeks. It showed less redness and better wound appearance.

Blinding was low risk in only about half the trials. One review's wound table is mostly rodent and rabbit studies and cell lines, plus a single human trial.

A randomized trial of a 7% centella cream at skin-graft donor sites ran 12 weeks against placebo, but we haven't seen its results. Don't extrapolate these doses, extracts or concentrations to a cosmetic serum.

How does centella asiatica help acne, and is there any human evidence?

There is no human evidence that it treats acne. In the dermatology review, all the acne data came from lab work.

• A methanol extract showed low antibacterial activity against C. acnes at 15 mg/mL in a disk-diffusion assay.
• An herbal mixture containing centella had an MIC of 31.25 µg/mL, but the authors noted it did not support antibacterial activity against the acne bacterium.
• Purified madecassoside (500 µM) reduced IL-1β, TLR2 expression and NF-κB translocation in human monocyte cells stimulated with the bacterium. That points to an anti-inflammatory role, not an antibacterial one.

The only registry entry we found, a planned split-face mask study of about 30 people, lists acne, rosacea and melasma as conditions. It's a registration record, not a result.

Beware the conflation: the ECa 233 trial involved atrophic acne scars after laser, which is wound healing, not acne treatment. Centella may calm irritation around acne treatment, but that's inference. We found no data showing it reduces pimples.

Can centella asiatica whiten skin or fade dark marks?

No. We found no controlled human trial showing that topical centella lightens skin, fades melasma or fades post-inflammatory hyperpigmentation. Our searches for one returned nothing relevant.

• The oral Korean trial found a 0.7% brightness gain, statistically significant against placebo but called modest by the authors.
• The burn trial counted reduced pigmentation among its scar-scale outcomes. That's healing in burn wounds, not cosmetic whitening.
• The pigment material in the dermatology review is a vitiligo cell study in which madecassoside protected human melanocytes from damage. That's the opposite of whitening.

A registered serum study of a centella extracellular-vesicle product lists "melanin hyperpigmentation" and "skin whitening" among its keywords. It is single-arm, unmasked and has 20 participants, and we haven't seen results.

If a centella product does leave skin looking brighter, the likely routes are indirect: less redness, a better barrier, faster recovery after procedures. That's speculation, not a finding.

Who is centella most likely to be useful for, and who can skip it?

Dry, irritated or recovering skin is the best-supported fit; for other skin types the evidence does not show a clear benefit. This is our synthesis, not a quoted conclusion.

• Dry or irritated skin that wants a gentle hydrating product. The evidence is small hydration and TEWL studies.
• Skin recovering from procedures such as lasers. The evidence is small RCTs, some of them multi-ingredient products.

• Anyone buying it to treat acne or fade dark spots.
• Anyone already happy with a well-formulated moisturizer.

Why do brands pair centella with other ingredients, and which one is doing the work?

Pairings are sensible formulation, but they make it hard to credit centella. A double-blind RCT [citation needed] of a panthenol-enriched mask for recovery after facial laser contained centella extract and bisabolol, and its authors credit the ingredients collectively. The batik trial compared centella with ceramide, not plain vehicle.

Ceramides and panthenol have their own barrier and soothing evidence, and the studies we retrieved don't isolate centella's share. The cleanest design we found was ECa 233 gel against placebo: one standardized extract, one control.

• A vehicle-only control arm.
• A named, standardized extract, such as ECa 233.
• A stated concentration. Studies used 0.05%, 2.5–5% and 7%; most products don't disclose theirs.
• A single-ingredient comparison.

What we still don't know

• No meta-analysis exists for cosmetic or barrier outcomes. The only systematic review we found covers wound healing, with four trials mixing oral and topical use.
• Cosmetic studies are small (25 to 30 people) and short (4 weeks), and some are industry-linked.
• Extract type, standardization and concentration vary, so results can't be pooled.
• We found no long-term data and no head-to-head comparison with niacinamide, azelaic acid, retinoids or ceramide moisturizers.