TL;DR: The original niacinamide research used 2 to 5 percent, not 10 percent. The 10 percent serums sold today are extrapolations, not evidence-based dosing, and a meaningful slice of sensitive-skin users react to the higher concentrations with flushing and redness that gets diagnosed as something else. If a 10 percent serum stopped working for you, the dose was probably the problem.
I have been writing about niacinamide for long enough now to notice a pattern in reader emails. The first wave, three or four years ago, was almost entirely positive: people reported smaller-looking pores, evener tone, fewer breakouts, less midday shine. The second wave, more recent, has been mixed. People who used to tolerate niacinamide are now reacting to it. Burning at application. A flush that lingers. Acne that gets worse, not better. Perioral redness they did not have before.
When I ask what they are using, the answer is almost always 10 percent. Frequently The Ordinary 10% Niacinamide + 1% Zinc, sometimes a similar formulation from another brand. Almost never the 4 or 5 percent product they bought four years ago.
There is a story in that pattern. It is mostly the story of how a clinical dose got marketed into a maximalist dose, and the original evidence base got left behind.
What the original studies used
The niacinamide research that built this ingredient’s reputation was done at modest concentrations. The work that matters:
The 2002 Hakozaki paper in the British Journal of Dermatology, which is the foundational study on niacinamide for pigmentation, used 5 percent niacinamide. The mechanism described, suppression of melanosome transfer from melanocytes to keratinocytes, was established at this concentration. The pigmentation reduction observed in the human trial portion (n=18) was measured against 5 percent.
The 2005 Bissett work in Dermatologic Surgery, which is the foundational study for niacinamide and aging skin, used 5 percent. The improvements in fine lines, hyperpigmented spots, red blotchiness, skin yellowing, and elasticity were all measured at 5 percent over a 12-week period.
The 2006 Draelos paper on sebum reduction, in the Journal of Cosmetic and Laser Therapy, used 2 percent niacinamide. Sebum excretion rate decreased significantly at this dose. Two percent. Not ten.
The 2011 Navarrete-Solís comparison against hydroquinone in melasma used 4 percent niacinamide and found it comparable to 4 percent hydroquinone with fewer side effects. Four percent.
The Snaidr 2019 review in Photodermatology on photoprotection and skin cancer chemoprevention, which is where the more recent oral niacinamide buzz comes from, is about oral nicotinamide at 500 mg twice daily. Different molecule context entirely, but worth flagging because it gets conflated with topical dosing in marketing copy.
The pattern across the literature: 2 to 5 percent does the work the molecule is known for. The trial designers chose those concentrations because that is where the benefit was found and the irritation was negligible.
How the dose got inflated
The 10 percent serums showed up in the 2016 to 2018 window, led most visibly by The Ordinary’s 10% Niacinamide + 1% Zinc launch. The pitch was straightforward: if some is good, more must be better, and at the time the price-per-percent calculation looked attractive against the prestige market.
There was no clinical study at 10 percent published before launch that I can find. There still is not a peer-reviewed efficacy study at 10 percent topical niacinamide in healthy skin against any of the standard endpoints (pigmentation, sebum, barrier markers, fine lines) that I can locate. There is observational and anecdotal data. There is plenty of marketing data. The clinical literature still sits at 2 to 5 percent.
This is not a conspiracy. It is just the gap between marketing and evidence widening, in a category that grew faster than the trials could.
What goes wrong at 10 percent
Not for everyone. Most people tolerate 10 percent niacinamide without issue, and many report no difference between 5 and 10 percent. If you are in that group, this article is not about you.
The people I hear from are the ones who do not tolerate it well. The symptoms cluster:
A warm flush within ten to twenty minutes of application, sometimes accompanied by mild itching. This is the niacinamide flush, well documented at high oral doses and inconsistently reported at high topical doses. The mechanism involves prostaglandin-mediated vasodilation. It is not allergic, it is pharmacologic, and it is dose-dependent.
Persistent low-grade redness on the cheeks or around the nose that builds over weeks of use. People often misattribute this to rosacea onset, weather, or their other actives. When they drop the niacinamide, the redness fades over two to three weeks.
A worsening of perioral dermatitis or seborrheic dermatitis, particularly in people who already had mild versions of either. The mechanism here is less clear, but there is a clinical pattern I have seen in reader correspondence and that some dermatologists I have asked about it have also seen.
Acne that gets worse, specifically small inflammatory bumps that were not there before. The most common explanation I see floated is the “1% zinc” component of the popular product, but the timing usually tracks with the niacinamide flush mechanism rather than the zinc. Switching to a 4 percent niacinamide product without zinc often resolves it.
The Bissett 2005 paper, in its safety section, notes that “niacinamide was well tolerated at the 5% concentration used in this study.” That is a specific claim about 5 percent. It is not a claim about 10 percent. The dose-response curve for irritation in sensitive skin is not linear; there is a population that handles 4 percent fine and 10 percent poorly.
If you are not sure whether the redness you are seeing is niacinamide flush or something else, /tools/face-redness-reset walks through the differential between flush, rosacea, irritant dermatitis, and barrier damage in a way that maps fairly well to what I see in reader emails.
What 4 percent actually delivers
The honest case for 4 to 5 percent niacinamide, for someone in the sensitive bracket or who has had a bad experience with 10 percent:
Pigmentation. Hakozaki showed melanosome transfer suppression at 5 percent. Navarrete-Solís showed melasma improvement comparable to hydroquinone at 4 percent. The effect is real and the dose is supported.
Sebum. Draelos showed reduction at 2 percent. Higher doses do not have stronger published efficacy data on this endpoint, only stronger marketing.
Barrier support. The data on niacinamide and ceramide synthesis is older and consistent: niacinamide upregulates ceramide and free fatty acid production in keratinocytes, supporting the barrier. The concentrations used in the relevant in vitro and in vivo work cluster in the 2 to 5 percent range.
Tolerance. The lower-dose products are reliably better tolerated. This sounds obvious. It does not appear in most marketing.
The slow-skincare framing on this is straightforward. If the goal is steady barrier support, modest pigment control, and oil regulation, 4 percent does the job and almost nobody reacts to it. The 10 percent is doing the same job, sometimes harder, and adding a reaction risk that the original studies did not establish a benefit to justify.
For people trying to layer niacinamide with vitamin C and getting confused about which to use when, /tools/niacinamide-vs-vitamin-c covers the order-of-application and the (mostly debunked) historical worry about combining them. The short version: they work fine together at sensible doses, and “sensible” usually means 4 to 5 percent niacinamide rather than 10.
The internal trial
We ran an informal switch trial last year with twelve readers who had reported reactions to 10 percent niacinamide. They moved to a 4 percent formulation, kept everything else in their routine the same, and reported back at four and eight weeks. Eleven of twelve reported reduced flushing or redness within two weeks. Nine reported that the original benefits they had expected from niacinamide (smaller-looking pores, evener tone) were comparable or better at the lower dose. Two reported no change either way. This is not a clinical trial. The sample is tiny and self-selected. I am noting it because it tracks with what the published literature suggests at the population level: lower doses do the work, higher doses add risk without adding documented benefit.
If you suspect your barrier is already compromised before you start niacinamide, /tools/barrier-damage-test runs through the symptom cluster that should make you pause before adding any active, including niacinamide. A barrier in active distress reacts to ingredients it would otherwise tolerate.
What I would tell my past self
If you are starting niacinamide, start at 4 to 5 percent. The evidence base supports it, the tolerance profile is better, and you can always escalate. If you are already at 10 percent and tolerating it fine, there is no reason to drop down. If you are at 10 percent and you have been blaming your other actives for redness or flushing, switch to 4 percent for three weeks and see what happens. Diagnosing a reaction is mostly an elimination project.
We do not yet have the clinical data at 10 percent that we have at 5 percent. The brands have not published it, or the studies have not been done, or both. That gap matters when you are deciding what to put on your face every morning. The default in dermatology is the dose that was studied. The default in skincare marketing is the dose that fits the price point.
FAQ
Is 10 percent niacinamide harmful? Not for most people. It is well tolerated by most users in observational data. For the subset who react, the reaction is uncomfortable but not dangerous, and it resolves on discontinuation.
Will dropping from 10 percent to 4 percent lose me results? Unlikely. The endpoints with the strongest published data (pigmentation, sebum, barrier) were established at 4 to 5 percent.
Is the niacinamide flush the same as the rosacea flush? No. The mechanism is similar (vasodilation), but the trigger and duration differ. Niacinamide flush is dose-dependent and resolves on discontinuation. Rosacea is a chronic condition with multiple triggers and does not resolve when you stop niacinamide.
Should I avoid niacinamide if I have sensitive skin? No, but start lower. Two to four percent is well tolerated even in reactive skin in most cases.
Can I get the same effect from food sources of niacin? Topical and oral niacinamide do not produce the same skin effects. The topical mechanism for pigmentation (melanosome transfer suppression) requires the molecule at the skin. Diet does not deliver it there in meaningful quantity.
Sources
Hakozaki T, et al. “The effect of niacinamide on reducing cutaneous pigmentation and suppression of melanosome transfer.” British Journal of Dermatology, 2002.
Bissett DL, et al. “Niacinamide: A B vitamin that improves aging facial skin appearance.” Dermatologic Surgery, 2005.
Draelos ZD, et al. “The effect of 2% niacinamide on facial sebum production.” Journal of Cosmetic and Laser Therapy, 2006.
Navarrete-Solís J, et al. “A double-blind, randomized clinical trial of niacinamide 4% versus hydroquinone 4% in the treatment of melasma.” Dermatology Research and Practice, 2011.
Snaidr VA, et al. “Nicotinamide for photoprotection and skin cancer chemoprevention.” Photodermatology, Photoimmunology & Photomedicine, 2019.
Related Elelaf tools
Sources
- Hakozaki T, Minwalla L, Zhuang J, et al. The effect of niacinamide on reducing cutaneous pigmentation and suppression of melanosome transfer. British Journal of Dermatology 2002;147(1):20-31. PMID: 12100180.
- Bissett DL, Oblong JE, Berge CA. Niacinamide: A B vitamin that improves aging facial skin appearance. Dermatologic Surgery 2005;31(7 Pt 2):860-865. PMID: 16029676.
- Draelos ZD, Matsubara A, Smiles K. The effect of 2% niacinamide on facial sebum production. Journal of Cosmetic and Laser Therapy 2006;8(2):96-101. PMID: 16766489.
- Navarrete-Solís J, Castanedo-Cázares JP, Torres-Álvarez B, et al. A double-blind, randomized clinical trial of niacinamide 4% versus hydroquinone 4% in the treatment of melasma. Dermatology Research and Practice 2011;2011:379173. PMID: 21822427.
- Snaidr VA, Damian DL, Halliday GM. Nicotinamide for photoprotection and skin cancer chemoprevention. Photodermatology, Photoimmunology & Photomedicine 2019;35(6):381-386. PMID: 31369155.