Topical Tranexamic Acid: Does It Reach Your Pigment, or Just Sit on Top?
Probably some of it reaches, but less than the label implies. Tranexamic acid is small (157 daltons) and strongly water-loving, which is exactly the profile the stratum corneum resists. In trials, plain topical tranexamic acid underperforms the same drug taken orally or pushed in with a needle. Treat it as an adjunct, not a monotherapy.
Every autumn I see the same thing. Someone arrives in Hongdae after a hot summer elsewhere, having bought three tranexamic acid serums in one afternoon and layered all three because the percentages differed and higher felt safer. What I find is rarely more even skin — it is a tight face that stings at the cheekbones, and pigment that has not moved.
What is tranexamic acid, and what is it actually for?
Tranexamic acid is a synthetic lysine analogue that stops plasminogen converting to plasmin, and it exists as a medicine because that stops bleeding. In Korea that is still its legal identity: a 250 mg oral capsule sold here is classified 전문의약품, prescription-only, approved for abnormal bleeding from fibrinolytic hyperactivity, at a usual adult dose of 750 to 2,000 mg a day in divided doses. Melasma is not on that label.
The pigment story is a side effect that became a product category. Plasmin sits upstream of inflammatory signals that nudge melanocytes, so inhibiting it may calm pigment driven by inflammation and UV. That is the mechanism every serum is selling — and a plausible mechanism is not a proven topical result.
How much tranexamic acid is in your serum versus in the studies?
Cosmetic tranexamic acid is commonly marketed at 3% to 5%. The clinical literature uses a startling spread for the same claim: a 2024 systematic review in Cutis pooling 12 trials lists topical doses of 0.4%, 0.5%, 1.8% liposomal, 4%, 5% and 10% — a twenty-five-fold range.
Tested head to head, concentration mattered less than you would expect. A double-blind randomised trial in Universa Medicina compared topical 10% against 5% in 16 women with epidermal melasma, eight per arm, twice daily for eight weeks, and found no statistically significant difference in MASI or pigmentation score. Eight people per arm cannot establish a dose-response curve, or rule one out — but a bigger number on the box is not buying a demonstrated difference.
Can a water-loving molecule get through your stratum corneum?
Partly, and nobody has measured it well in you. Tranexamic acid weighs 157 daltons, small enough to cross a healthy barrier in principle. The problem is the rest of the profile: water solubility around 167 mg/mL and a predicted log P of about −0.09 at pH 5. The stratum corneum is a lipid wall, and a molecule that would rather stay in water does exactly that — sitting in the surface aqueous phase instead of partitioning in.
An in vitro study using MelanoDerm, a three-dimensional reconstructed human skin model, ran free and encapsulated tranexamic acid for 21 days. Free 3% tranexamic acid brought relative melanin content to 83.9% of blank, missing significance (p = 0.0559). A PLGA-encapsulated formulation carrying only 0.5% reached 75.71% (p = 0.0118); a liposomal one carrying 0.25% reached 76.88% (p = 0.0186). Six times less drug, better result — delivery, not concentration, was the bottleneck. This was tissue in a dish, and the encapsulated systems were not loaded to matched concentrations.
The clinical trials say the same from the other end: the studies working hardest to show an effect bypass the barrier — microneedling, iontophoresis, injection. A 2024 review in Cosmetics states that intralesional delivery gives better availability in skin than topical, and that topical tranexamic acid is not suitable as monotherapy.
Why does the same serum behave differently around your eyes?
Nobody in this literature measures the eyelid. The scoring tool these trials use divides the face into forehead, cheeks and chin; the periorbital area, upper lip and neck are not scored. So "effective for melasma" is a finding about cheek and forehead skin.
Thinner skin is not a shortcut to a better result. Less barrier means less buffer: solvents, surfactants and preservatives arrive alongside the active, and what people notice around the eyes and mouth first is usually a dry, crepey sting rather than faded pigment. Periorbital darkness is also often vascular shadow or a hollow — and no amount of plasmin inhibition changes a shadow.
What is buyable here is unremarkable, which is fine. Korean pharmacies and cosmetics shops stock tranexamic acid mostly as thin aqueous ampoules and serums in the 2% to 5% band, usually paired with niacinamide; the pharmacy counter holds the oral drug, which needs a doctor.
Know your own skin before you buy anything
Actives belong one at a time, at the pace your barrier can take, and the first question is what state your skin is in today — not which ingredient has the best research. There is a version of self-care that is really just effort. Someone who is genuinely unwell does not recover by training harder; the body in that state needs rest, and pushing through turns a setback into an injury. Skin is the same. A face that is stinging, flaking or reacting to products it used to tolerate is not under-treated — it is asking to be left alone.
When the barrier is compromised, the first move is subtraction. Strip back to a gentle cleanser, one bland moisturiser and sunscreen; hold until the stinging stops; then add a single active and wait two weeks before judging it. Pigment is patient. None of this is a diagnosis or treatment plan, and pigmentation that persists or changes should be seen by a dermatologist.
The case against some of this
I implied tranexamic acid reaches melanocytes when you rub it on. The strongest reading of the data does not support that. In the MelanoDerm study above, free 3% tranexamic acid missed significance outright (p = 0.0559); only the encapsulated versions cleared it. A reconstructed tissue model has no circulation and no sebum film, so it is weak evidence about faces either way.
I treated the concentration question as settled. It is not, in either direction. The 10%-versus-5% trial had eight people per arm and its abstract reports no within-group statistics, so "both worked" rests on full text I could not read. The Cutis review is blunter: small samples, high dropout, subjective mMASI scoring, and formulations varied enough to make pooling questionable.
I leaned on topical tranexamic acid roughly matching hydroquinone. That is not uniformly true. The Cosmetics review reports 5% cream producing MASI reductions similar to 2–4% hydroquinone at 12 weeks, but the Cutis review notes Xing and colleagues found liposomal tranexamic acid alone improved melanin index less than hydroquinone. And the best topical trial I could read in full — randomised, double-blind, placebo-controlled, at West China Hospital — tested 3% tranexamic acid essence with iontophoresis against vehicle with iontophoresis in 30 women, analysing 26. MASI change at 12 weeks favoured tranexamic acid (p = 0.044), but the rate-of-decline comparison missed significance (p = 0.069), inter-rater agreement was low to moderate (ICC 0.19–0.62), and there was no tranexamic-acid-alone arm.
Narrowing, not retracting: topical tranexamic acid is still worth using — as a once-daily adjunct to sunscreen, on an intact barrier, for inflammation-driven pigment, judged at three months, and not as a replacement for oral therapy or a procedure.
Is 10% tranexamic acid better than 5%?
Not on the available evidence. The one head-to-head randomised trial found no significant difference over eight weeks, though eight people per arm could not have detected a modest one. Delivery system appears to matter more than percentage.
Can I buy tranexamic acid tablets for dark spots in Seoul?
Not over the counter. Oral tranexamic acid is prescription-only in Korea and approved for bleeding disorders, so pigment use is off-label and needs a doctor who can review your clotting history first.
Should I use tranexamic acid around my eyes?
No trial has scored that area, and thin periorbital skin tends to react to the vehicle before it responds to the active. Dark circles there are more often vascular or structural than pigmentary.
If you cannot tell whether your pigment is inflammatory, vascular or structural — or whether your barrier can take an active right now — that is the judgement we make in person at Laurel Studio in Hongdae. One consultation beats three more serums.
Sources
- Skin Lightening Efficacy of Tranexamic Acid in Topical Delivery Systems — in vitro, MelanoDerm 3D reconstructed human skin model, 21 days; molecular weight 157 Da, log P −0.09, solubility 167 mg/mL; free 3% TXA p = 0.0559. Full text read.
- Chowdhary et al., oral tranexamic acid 250 mg once daily versus 500 mg twice daily, Archives of Dermatological Research — open-label comparative study in humans, 132 enrolled, no control arm, 16 weeks. Full text read.
- Korean drug information, oral tranexamic acid 250 mg capsule — classified 전문의약품 (prescription-only), bleeding indications, adult dose 750–2,000 mg/day.
- Comparative efficacy of topical 10% versus 5% tranexamic acid in women with melasma, Universa Medicina — double-blind RCT, 16 women, 8 per arm, twice daily, 8 weeks. Abstract only.
- Parać and Bukvić Mokos, "Unmasking Melasma: Confronting the Treatment Challenges", Cosmetics, 2024 — narrative review; topical TXA not suitable as monotherapy; photoprotection essential regardless of active.
Sources for the case-against section
- Olugbade and Negbenebor, "The Use of Tranexamic Acid and Microneedling in the Treatment of Melasma: A Systematic Review", Cutis, 2024 — 12 randomised and clinical trials, mostly small split-face designs; small samples, high dropout, subjective mMASI scoring, heterogeneous formulations; Xing et al. found topical liposomal TXA alone inferior to hydroquinone. Full text read.
- Treatment of Melasma with Tranexamic Acid Essence Combined with Iontophoresis: A Randomized, Double-Blind, Placebo-Controlled Clinical Trial — 30 women randomised, 26 analysed, 3% TXA essence plus iontophoresis versus vehicle plus iontophoresis, twice weekly for 3 months; rate-of-decline p = 0.069, MASI ICC 0.19–0.62, no TXA-alone arm. Full text read.
- Parać and Bukvić Mokos, Cosmetics, 2024 — cited again for the statement that topical tranexamic acid is less effective than oral or injectable routes.
