Face-off
Niacinamide vs azelaic acid vs tranexamic acid
The three most talked-about brighteners act on different steps of the pigment pathway, and each has been tested on different problems. Here is what the trials show in melasma, acne marks and rosacea, and where each one earns its place.
Revelika editors10 min read
Ask about dark spots and someone will recommend one of three ingredients: niacinamide, azelaic acid or tranexamic acid. They are often sold as interchangeable “brighteners”. They aren’t. Each has its own mechanism, its own evidence and its own tolerability, and, most importantly, each was studied in different conditions. A result in melasma does not carry over to acne marks, or the other way round, so below we name the condition every time.
, ,
Tap an underlined ingredient to see what it does.
Niacinamide: slows pigment delivery
Niacinamide doesn’t stop melanocytes from making melanin. In the classic study by Hakozaki and colleagues it had no effect on tyrosinase activity or on melanin production in cultured melanocytes. What it did was cut the transfer of melanosomes, the packets of pigment, into neighbouring keratinocytes by 35–68% 1. Pigment is still made; less of it reaches the surface.
In the same paper, a 5% niacinamide moisturiser reduced hyperpigmentation more than its vehicle after four weeks in 18 Japanese women 1. The study is small and came from a cosmetics manufacturer’s research lab. That is no reason to dismiss it, but it is a reason not to oversell it.
For melasma, the main evidence is a double-blind trial by Navarrete-Solís and colleagues: 27 patients applied 4% niacinamide to one side of the face and 4% hydroquinone to the other for eight weeks, with sunscreen 2. Colorimetry found no difference between the sides. Good-to-excellent improvement was seen in 44% with niacinamide and 55% with hydroquinone; side effects in 18% and 29%. In short, niacinamide is slightly weaker but gentler, and that rests on one small trial.
In rosacea, the niacinamide data are about the barrier, not pigment. In a four-week study of 50 people, a niacinamide moisturiser improved barrier function and stratum corneum hydration 3. That supports niacinamide as supportive care, not as a rosacea treatment.
Revelika’s data puts niacinamide’s effective strength at 5% and our upper limit at 10%. The trials above used 4–5% 1,2, so 15–20% serums are not backed by them. Our data records niacinamide as compatible with pregnancy.
Azelaic acid: the most tested of the three
Azelaic acid is the only contender that exists as an approved drug with formal indications. In the US, a 20% cream is approved for mild-to-moderate inflammatory acne 8 and a 15% gel for the papules and pustules of mild-to-moderate rosacea 7. Its labels say plainly that the exact mechanism is unknown; in vitro it is active against acne-associated bacteria 8. The usual explanation for its effect on pigment, a selective action on overactive melanocytes, is covered in our ingredient profile.
In melasma, azelaic acid has the largest evidence base of the three. In a double-blind trial by Baliña and Graupe, 329 women used 20% azelaic acid or 4% hydroquinone for 24 weeks; there were no significant differences between groups, and 65% on azelaic acid had good or excellent results 4. An earlier trial of 155 patients found 20% azelaic acid ahead of 2% hydroquinone: 73% good-to-excellent results against 19% 5. Both studies are more than 30 years old, and in both, participants used sunscreen.
For darker skin (phototypes IV–VI) there is a vehicle-controlled trial: over 24 weeks, the 20% cream reduced facial hyperpigmentation significantly more than vehicle, at the cost of slightly more burning and stinging early on 6. For acne marks, 20% azelaic acid matched 5% tranexamic acid in a randomised trial 14; more on that below.
In rosacea the evidence is strong, with a caveat. The 15% gel is approved for papules and pustules, and efficacy for redness alone, without bumps, has not been evaluated; the label says so 7.
The price is irritation. In trials of the 15% gel, the most common reactions were burning, stinging or tingling (29%), itching (11%) and scaling or dryness (8%), usually in the first weeks 7. There are isolated reports of hypopigmentation, so people with darker skin are advised to watch for early signs 7.
Revelika lists 15% as the effective strength and 20% as the maximum, exactly the strengths of the approved products. If a product contains less, the 15–20% trials don’t directly apply to it. Our data records azelaic acid as compatible with pregnancy; the gel’s label notes that it is minimally absorbed through skin and maternal use is not expected to expose the fetus 7.
Tranexamic acid: promising, but uneven data
Tranexamic acid is an anti-bleeding drug: it stops plasminogen from being converted to plasmin. Plasminogen is also found in the basal cells of the epidermis, which is why researchers expected that blocking this pathway might quiet the signals that spur melanocytes 10. In an open study by Kim and colleagues, 23 people with melasma applied 2% tranexamic acid for 12 weeks: mMASI scores improved in 22 of 23, biopsies showed less epidermal melanin and lower endothelin-1, and vessel counts tended to fall 9. That suggests it acts on melasma’s vascular side as well. But there was no control group.
The most rigorous topical study is a double-blind split-face trial (Kanechorn Na Ayuthaya, 2012): 23 women applied 5% tranexamic acid to one side and the plain vehicle to the other for 12 weeks. Both sides lightened, but the active did no better than the vehicle. The tranexamic acid side also had more redness 11.
Comparisons with hydroquinone look better. In a randomised trial of 100 patients in India, a 5% tranexamic acid solution cut MASI by 27% over 12 weeks and 3% hydroquinone by 26.7%; satisfaction was higher with tranexamic acid because of fewer side effects 12. In a double-blind Iranian trial of 60 women, 5% tranexamic acid and 2% hydroquinone lowered MASI with no significant difference between groups 13. A 2017 review of the clinical studies still concluded that its efficacy in melasma is not adequately established 10.
For acne marks there is one randomised trial of 60 people: 5% tranexamic acid and 20% azelaic acid gave comparable results over 12 weeks, with more side effects on azelaic acid at week four 14.
Tablets are a separate story. Oral tranexamic acid has been studied more than the topical form, mostly in Asian patients, and a 2018 review calls it an effective option for stubborn melasma once contraindications have been carefully screened for 15. It is a prescription decision, not skincare. In the US the tablets are approved for heavy menstrual bleeding; the label lists active thrombosis, a history of thrombosis and an intrinsic risk of clotting as contraindications, and carries a separate warning about combined hormonal contraception 16.
Revelika lists 3% as the effective topical strength and 5% as the maximum; trials used 2% and 5% 9,11,12,13,14. Our data marks it “with caution” in pregnancy. The cited studies contain no data on topical use in pregnancy, and the tablet label notes that tranexamic acid crosses the placenta and has not been adequately studied in pregnant women 16.
| Niacinamide | Azelaic acid | Tranexamic acid | |
|---|---|---|---|
| How it works | Slows melanosome transfer to keratinocytes 1 | Exact mechanism unknown; active against acne-associated bacteria 8 | Blocks the plasminogen–plasmin pathway; lowers endothelin-1 9,10 |
| Best evidence for | Gentle fading of spots; melasma in one small trial 1,2 | Acne and papulopustular rosacea (approved drugs); melasma vs hydroquinone 4,7,8 | Melasma vs hydroquinone; tablets by prescription only 12,13,15 |
| Effective % (Revelika) | 5% | 15% | 3% |
| Maximum % (Revelika) | 10% | 20% | 5% |
| Pregnancy (Revelika) | Compatible | Compatible | With caution |
| Typical irritation | Low: side effects in 18% vs 29% on hydroquinone 2 | Burning or tingling in 29%, mostly early on 7 | Usually mild; redness in one trial 11 |
Limited evidence
One double-blind split-face trial, 27 people, 8 weeks, against 4% hydroquinone.2
Limited evidence
Small vehicle-controlled studies of a 5% cream over 4 weeks.1
Limited evidence
Barrier support in a moisturiser only; not a treatment.3
Moderate evidence
Two large double-blind trials of 20% cream against hydroquinone; both over 30 years old.4,5
Azelaic acid · PIH and darker skin
Moderate evidence
Vehicle-controlled trial in phototypes IV–VI; head-to-head with tranexamic acid for acne marks.6,14
Strong evidence
15% gel approved for papules and pustules; redness alone not evaluated.7
Strong evidence
20% cream approved for mild-to-moderate inflammatory acne.8
Tranexamic acid (topical) · melasma
Limited evidence
Matches hydroquinone in small trials, but no better than vehicle in the only vehicle-controlled one.9,11,12,13
Tranexamic acid (topical) · acne PIH
Limited evidence
One randomised trial: comparable to 20% azelaic acid.14
Tranexamic acid (oral) · melasma
Moderate evidence
Better studied than topical; prescription drug with contraindications.15,16
Not enough evidence
No controlled trials among our sources.
Which one for you
The short answer: azelaic acid for acne, acne marks and bumpy rosacea; tranexamic acid for melasma when gentleness matters; niacinamide as the best-tolerated and the most modest. In more detail:
- Melasma. Azelaic acid at 15–20% is the best studied 4,5; topical tranexamic acid is usually gentler, but its data are less consistent 11,12,13. In all these trials, sunscreen was part of the regimen 2,4,5,11. Niacinamide is the gentle option for skin that won’t tolerate acids 2. Stubborn melasma is a reason to see a dermatologist: tranexamic acid tablets are for a doctor to prescribe 15,16.
- Acne marks. Azelaic acid, because it treats both the marks and the acne behind them 8,14. Tranexamic acid is an alternative with a gentler start 14.
- Rosacea with papules and pustules. Azelaic acid 15% 7. Niacinamide in a moisturiser, for the barrier 3. There isn’t enough evidence for tranexamic acid in rosacea.
- Sensitive skin, or your first active. Niacinamide at 4–5% 2.
- Pregnancy. Revelika’s data records niacinamide and azelaic acid as compatible and tranexamic acid as “with caution”. Talk your choice through with your doctor.
- Darker skin. Azelaic acid has been studied in phototypes IV–VI 6, but watch for signs of hypopigmentation 7.
Can you combine them?
Revelika’s data lists no conflicts between them: the field is empty for all three. Azelaic acid’s profile lists niacinamide and tranexamic acid as synergies, and tranexamic acid’s lists niacinamide and azelaic acid. Niacinamide’s own synergies are different: hyaluronic acid, zinc oxide and ceramides.
Little has been tested in the clinic. In a 2025 randomised double-blind trial of 99 melasma patients, a cream with 5% tranexamic acid and 4% niacinamide, and a niosomal version at 2% + 2%, worked as well as 4% hydroquinone over three months with far fewer adverse reactions: about 3% against 18% 17. But there were no niacinamide-only or tranexamic-only groups, so the trial cannot show that the pair beats either ingredient alone. Among our sources there are no controlled trials of combinations with azelaic acid.
The practical upshot: niacinamide with tranexamic acid is a sensible pairing with one good trial behind it. Add azelaic acid gradually and watch how your skin responds, since it is the most irritating of the three 7. And in the melasma trials, daily SPF was a fixed part of the regimen 2,4,5,11: without it, there is nothing to compare.
More from this issue
Dossier9 min read
Red marks, brown marks: what actually fades them
One is pigment, the other blood vessels. Why post-acne marks need different fixes, and how good the evidence for each really is.
Explainer9 min read
Homosalate and sunscreen filters: what the EU changed
The EU narrowed homosalate to face products at 7.34%. Why, what the FDA says, and what to do with your bottle.