SkincareIngredient GuideAug 2026

Niacinamide: What It Does, What Strength, and What the Evidence Shows

Niacinamide is a form of vitamin B3, and it turns up on a lot of ingredient lists. The honest version is more interesting than the marketing: a handful of small human trials found real changes in the appearance of skin, most of the mechanism comes from cell culture rather than from faces, and the strength printed on the front of the bottle is often higher than anything that has been tested on its own. Here is what each study actually did, and what it did not.

2 to 5%

Strengths used in the skin trials

n = 50

Size of the photoaging trials

12 wk

How long those ran

What Niacinamide Is

Niacinamide is one of the two everyday forms of vitamin B3. The other is nicotinic acid, which most labels call niacin. Same vitamin, two different molecules, and they behave differently in the body. A 2020 review in Skin Therapy Letter opens by pointing out that nicotinamide is frequently confused with its precursor nicotinic acid, which is exactly the confusion that produces most of the internet arguments about this ingredient (Huber and Wong, PMID 33196157).

The practical version: if a label says niacinamide, or nicotinamide, or vitamin B3 amide, those are the same molecule. If it says niacin or nicotinic acid, that is the other one. The flushing reaction people associate with B3 supplements, the hot red face, is attributed to nicotinic acid acting at a receptor called GPR109A, which triggers prostaglandin release from Langerhans cells in the skin (Stern, 2007, PMID 21291680). That is a description of nicotinic acid, not of niacinamide.

Mechanistically, a 2014 review in Skin Pharmacology and Physiology describes niacinamide as a water-loving substance the body already makes, with effects that depend on concentration, working in part by inhibiting PARP-1 and so influencing NF-kB mediated transcription. That same review is careful about how far the clinical evidence goes: its authors concluded the existing data were not sufficient for a scientifically founded evaluation, while calling the prospects interesting (Wohlrab and Kreft, PMID 24993939). Both halves of that verdict matter.

What The Trials Measured

Four separate questions have been studied, and it helps to keep them apart, because a study of sebum tells you nothing about wrinkles.

Skin barrier and ceramides

The barrier evidence starts with a 2000 paper that is largely laboratory work. In cultured human keratinocytes, six days of nicotinamide at 1 to 30 micromol per litre raised ceramide synthesis 4.1 to 5.5 fold, glucosylceramide 7.4 fold and sphingomyelin 3.1 fold. Those are cell-culture numbers, not skin numbers, and they should not be read as what happens on a face. The same paper also reports that topical application raised ceramide and free fatty acid levels in the stratum corneum and lowered transepidermal water loss in dry skin, although the abstract does not give the design or size of that human portion (Tanno et al., 2000, PMID 10971324).

Pigmentation and skin tone

The 2002 Hakozaki paper is more nuanced than the brightening claims it gets cited for. Niacinamide had no effect on the catalytic activity of mushroom tyrosinase and no effect on melanogenesis in cultured melanocytes. What it did do, in a keratinocyte and melanocyte coculture model, was inhibit melanosome transfer by 35 to 68 percent. Again, that is a dish, not a face. The human half of the paper ran two studies in Japanese women: 18 subjects with hyperpigmentation using 5 percent niacinamide moisturizer against vehicle in a paired design, and 120 subjects with facial tanning assigned to two of three treatments, vehicle, sunscreen, or 2 percent niacinamide plus sunscreen. Hyperpigmentation fell and skin lightness rose versus vehicle after four weeks (PMID 12100180). The population studied was Japanese women, and the paper comes from a Procter and Gamble laboratory, both of which belong in any honest summary. Pigmentation conditions with their own clinical definition, such as melasma, are a separate question and are covered in the melasma guide.

Oil and shine

Two trials of 2 percent niacinamide were run in parallel and they did not agree with each other, which is the interesting part. In Japan, 100 subjects were split into two groups of 50, one using a 2 percent niacinamide moisturizer and one a placebo moisturizer for four weeks: sebum excretion rate was significantly lower in the niacinamide group at both two and four weeks. In the United States, 30 subjects ran a six-week split-face study, and there the casual sebum level was significantly reduced but the sebum excretion rate was not. The authors phrased their own conclusion accordingly, that 2 percent niacinamide may lower sebum excretion rate in Japanese individuals and casual sebum level in Caucasian individuals (Draelos et al., 2006, PMID 16766489). Anyone quoting only the Japanese half is quoting half a paper.

The appearance of aging skin

Two 12-week, double-blind, split-face trials tested niacinamide against a vehicle for the appearance of aging skin. Both used 5 percent niacinamide against the identical moisturizer without it, in 50 white women, with the sides randomised left and right. The 2004 trial, in women aged 40 to 60, reported significant improvement versus control in fine lines and wrinkles, hyperpigmentation spots, texture, red blotchiness and skin yellowing (PMID 18492135). The 2005 trial reported the same picture apart from texture, plus improved elasticity measured by cutometry (PMID 16029679). Both were authored by researchers at Procter and Gamble.

A note on the rosacea study people cite: a 2005 study in Cutis enrolled 50 subjects with rosacea to test whether a niacinamide-containing facial moisturizer improved the stratum corneum barrier (PMID 16209160). It is worth knowing two things about it. It tested a whole moisturizer rather than niacinamide on its own, and the abstract available through PubMed describes the design and the measurements without reporting the numeric outcomes, so this page does not quote a result from it.

Which Percentage Was Actually Tested

This is the question most people are really asking, and the answer is that percentages on shelves and percentages in trials are two different sets of numbers. Here is what the human studies above used, and for what.

Strength	What it was tested for	Study
2%	Facial sebum, 4 and 6 weeks	Draelos 2006, PMID 16766489
2% with sunscreen	Facial tanning, Japanese women	Hakozaki 2002, PMID 12100180
4%	Inflammatory acne vs 1% clindamycin gel, 8 weeks	Shalita 1995, PMID 7657446
5%	Facial hyperpigmentation, 4 weeks	Hakozaki 2002, PMID 12100180
5%	Photoaging appearance, 12 weeks	Bissett 2004 and 2005, PMID 18492135, 16029679
10%	Inside combination or engineered products, not tested alone	PMID 39376788, PMID 30411631

Notice what is missing: no study in that list compared one strength against another. 5 percent beat a vehicle and 2 percent beat a vehicle, in different studies, on different endpoints, in different populations. That is not the same as 5 percent beating 2 percent, and no trial that ran that comparison turned up in the searches behind this page.

The 10 percent products deserve their own paragraph, because 10 percent is now a common marketing number. Searching PubMed on the title and abstract field for the phrases 10 percent niacinamide and 10 percent nicotinamide returned nothing for the first spelling and five records for the second, of which two involved human skin. One was a randomised, double-blind Brazilian trial of a product combining 10 percent nicotinamide with 5 percent magnesium ascorbyl phosphate and 5 percent hyaluronic acid, compared against 4 percent hydroquinone in women with facial melasma over 60 days. The two arms did not differ on the melasma severity index at day 14 or day 60 (p greater than 0.2), although colorimetric luminosity improved more with hydroquinone at day 60 (p equals 0.01), and the authors described the combined product as numerically inferior overall while noting it was safe and well tolerated (PMID 39376788). The other was a pharmaceutics paper in which a 10 percent nicotinamide adhesive extrudate was compared with a nicotinamide gel in acne patients, which compares two ways of delivering the same vitamin rather than 10 percent against a plain vehicle (PMID 30411631).

The honest reading: no published human comparison isolating what 10 percent adds over 5 percent turned up in these searches. That is an absence of evidence, not evidence that 10 percent is worse or useless. It simply means the number on the front of the bottle is a formulation choice, not a finding.

Niacinamide For Acne

A 1995 double-blind trial compared topical nicotinamide head to head with an antibiotic gel. It randomised 76 patients with moderate inflammatory acne, 38 to 4 percent nicotinamide gel and 38 to 1 percent clindamycin gel, applied twice daily for eight weeks. After eight weeks, 82 percent of the nicotinamide group and 68 percent of the clindamycin group were rated improved on physician global evaluation. Papule and pustule counts fell 60 percent versus 43 percent, and acne severity fell 52 percent versus 38 percent (Shalita et al., PMID 7657446).

Those numbers look decisive and they are not. None of the three differences reached statistical significance: the P values were 0.19, 0.168 and 0.161 respectively. The trial reports no detected difference between the two gels, which the authors framed as comparable efficacy. Just as importantly, there was no placebo arm. A trial with two active treatments and no vehicle control cannot tell you whether either one beat doing nothing over eight weeks, and acne fluctuates on its own.

The 2020 Cochrane review went looking for the same question across the whole randomised literature. Across 49 trials and 3,880 reported participants, run in clinics, hospitals, research centres and universities in Europe, Asia and the USA, the participants were mostly female, mostly aged 12 to 30, and mostly had mild to moderate acne. On nicotinamide specifically, the review found four studies comparing it against clindamycin or erythromycin, and none of them measured the review's treatment-response outcome, which was the participants' own global self-assessment of improvement. That left only tolerability comparable across studies (Liu et al., PMID 32356369):

Withdrawal for any reason, nicotinamide versus clindamycin: risk ratio 1.12, 95 percent CI 0.49 to 2.60, from 3 studies and 216 participants.
Withdrawal for any reason, nicotinamide versus erythromycin: risk ratio 1.40, 95 percent CI 0.46 to 4.22, from 1 study and 158 participants.
Total minor adverse events, nicotinamide versus clindamycin: risk ratio 1.20, 95 percent CI 0.73 to 1.99, from 3 studies and 216 participants.
All three were rated low-quality evidence, and every confidence interval crosses 1.

So the strongest claim the systematic evidence supports is that topical nicotinamide was tolerated about as well as the antibiotic gels it was compared with. The effectiveness question was not answered, because the trials did not collect the outcome the reviewers were looking for. That is a different statement from either "niacinamide clears acne" or "niacinamide does not work for acne", and anyone with acne that is scarring or not settling should be having this conversation with a dermatologist rather than with an ingredient list.

Oral Nicotinamide Is A Different Story

Search results for this vitamin mix two literatures that have almost nothing to do with each other. The largest randomised trials found for this page are not about serums at all. They are about swallowing nicotinamide to prevent skin cancers in people who have already had several, and it is worth understanding precisely because it is so often borrowed to sell a serum.

ONTRAC, Australia, 2015 (PMID 26488693)

A phase 3 double-blind trial randomly assigned 386 participants in a 1 to 1 ratio, so roughly 193 per arm, all of whom had had at least two non-melanoma skin cancers in the previous five years. They took 500 mg of nicotinamide or placebo twice daily for 12 months. New non-melanoma skin cancers were 23 percent lower in the nicotinamide group (95 percent CI 4 to 38, P equals 0.02). Basal cell carcinomas were 20 percent lower but the interval crossed zero (95 percent CI minus 6 to 39, P equals 0.12), and squamous cell carcinomas 30 percent lower (95 percent CI 0 to 51, P equals 0.05). Actinic keratoses were 11 to 20 percent lower across the checkpoints. Adverse events did not differ noticeably, and there was no evidence of benefit after the nicotinamide was stopped.

ONTRANS, Australia, 2023 (PMID 36856616)

The same design was run in solid organ transplant recipients with the same skin cancer history: 158 enrolled, 79 to nicotinamide and 79 to placebo, and the trial was stopped early because recruitment was poor. At 12 months there were 207 new keratinocyte cancers in the nicotinamide group and 210 in the placebo group, a rate ratio of 1.0 (95 percent CI 0.8 to 1.3, P equals 0.96). No benefit was seen on squamous cell or basal cell counts, actinic keratoses, or quality of life. The same regimen that worked in one population did nothing measurable in another.

A US Veterans Affairs cohort, 2025 (PMID 40960808)

A retrospective cohort, not a randomised trial, drawn from United States Veterans Affairs electronic health records, matched 12,287 patients prescribed nicotinamide 500 mg twice daily for more than 30 days against 21,479 unexposed patients. The exposed group had a mean age around 77, was 2 percent women and about 95 percent White, so it describes a narrow slice of one country. Overall skin cancer risk was 14 percent lower, rising to 54 percent lower when nicotinamide was started after a first skin cancer and falling away when started later. Among solid organ transplant recipients there was no significant overall reduction, although earlier use was associated with less cutaneous squamous cell carcinoma. The journal has published a correction to this article, so treat the exact figures as provisional.

Three things follow. First, every one of those results is oral nicotinamide, not a serum, and none of them says anything about what happens when you put niacinamide on your face. Second, all of them are in people who had already had at least one skin cancer, and the two randomised trials were restricted to people with a history of multiple skin cancers under dermatologist follow-up, so they do not generalise to a healthy person choosing a supplement. Third, ONTRAC and ONTRANS together are a live demonstration that a result in one population is not a result in every population. If oral nicotinamide is something you are weighing for skin cancer risk, that is a conversation for a dermatologist who knows your history.

Two Claims I Could Not Verify

Two things get repeated about niacinamide constantly. Neither one held up when I went looking for the human evidence behind it.

"Niacinamide and vitamin C cancel each other out"

I searched PubMed for a controlled human trial testing whether layering niacinamide with a vitamin C product reduces the effect of either one, and did not find one. So this page cannot confirm the claim, and cannot refute it either. What does exist is a randomised, double-blind trial in which 10 percent nicotinamide and 5 percent magnesium ascorbyl phosphate, a vitamin C derivative, were formulated in the same product and used twice daily for 60 days with no serious side effects reported (PMID 39376788). That shows the two can coexist in one bottle. It is not a test of layering two separate products, and it uses a derivative rather than pure ascorbic acid, so it settles less than it looks like it settles.

"Niacinamide causes flushing"

The flushing associated with vitamin B3 is documented for nicotinic acid, acting at the GPR109A receptor to trigger prostaglandin release (PMID 21291680). Nicotinic acid is the other form of the vitamin, and the published literature notes explicitly that the two are routinely confused with each other (PMID 33196157). In the sources read for this page, the flushing reaction is attributed to nicotinic acid rather than to niacinamide.

A third thing worth saying plainly: niacinamide does not have a clean, settled answer to the question "what percentage should I use". Where an ingredient page hands you a confident number, it is fair to ask which trial that number came from.

Where The Evidence Stops

The appearance trials are small and short: 50 participants, 12 weeks, split-face, and both were authored by researchers at Procter and Gamble. That affiliation is disclosed in the papers themselves, but it is still a commercial interest to weigh.
Much of the mechanism, the ceramide synthesis and the melanosome transfer numbers, comes from cell culture rather than from skin on a face. Cell-culture folds do not translate into face results.
The populations studied are narrow: white women aged 40 to 60 for photoaging, Japanese women for pigmentation and one of the two sebum arms, mostly female adolescents and young adults for acne.
The searches behind this page found no head-to-head comparison of 2 against 5 against 10 percent, and none comparing one commercial niacinamide serum with another.
The 2014 mechanism review concluded the data were not yet sufficient for a scientifically founded evaluation. No later topical trial found for this page is large enough to overturn that.

None of that makes niacinamide a bad ingredient. The trials above that reported on tolerability described it as well tolerated, the Cochrane review found no signal that it was tolerated worse than the antibiotic gels it was compared with, and several independent lines of human evidence point in the same direction. It does mean the confident, specific promises attached to it are running well ahead of the trials. If you have a skin condition with a name, rather than a preference about texture and tone, a dermatologist or pharmacist is the right person to talk to, and in Vietnam a pharmacist is easy to reach.

The Short Version

Niacinamide is vitamin B3 in its amide form. Nicotinamide is the same molecule. Niacin, or nicotinic acid, is the other form and the one linked to flushing.
Human trials used 2 percent for sebum, 4 percent for acne, and 5 percent for pigmentation and photoaging appearance. No trial comparing those strengths against each other was found.
10 percent is a common marketing number that has not been tested on its own against a vehicle in the papers found for this page.
For acne, a 1995 trial found no statistically significant difference between 4 percent nicotinamide gel and 1 percent clindamycin gel, and it had no placebo arm. Cochrane 2020 could not assess effectiveness because the nicotinamide trials did not measure the outcome it needed.
The 12-week photoaging trials reported real improvements in the appearance of lines, spots, texture, redness and sallowness, in 50 white women, and were run by the ingredient supplier.
The oral nicotinamide skin cancer trials are a separate literature about swallowed 500 mg doses in high-risk patients. They do not describe what a serum does.
The claim that niacinamide and vitamin C cancel each other out has no controlled human trial behind it that this page could find.

Frequently Asked Questions

What does niacinamide do for skin?+

In controlled human trials it has been measured against several different endpoints. Twelve-week split-face trials of 5 percent niacinamide in 50 white women with photoaging reported improvements versus the same moisturizer without it in fine lines and wrinkles, hyperpigmented spots, texture, red blotchiness and sallowness, plus elasticity by cutometry (Bissett et al., 2004, PMID 18492135; Bissett et al., 2005, PMID 16029679). Separate work reported reduced facial hyperpigmentation and increased skin lightness after four weeks in Japanese women (Hakozaki et al., 2002, PMID 12100180). These are small, short, industry-run trials, and the 2014 review of the mechanism concluded the overall data were not yet sufficient for a scientifically founded evaluation (Wohlrab and Kreft, PMID 24993939).

What percentage of niacinamide is best?+

The published skin trials used 2 percent for sebum, 4 percent for inflammatory acne, and 5 percent for pigmentation and photoaging appearance. No trial directly comparing those strengths against each other turned up in the searches behind this page, so no percentage has been shown to be best. A PubMed title and abstract search for 10 percent found two papers involving human skin, and in both the 10 percent was inside a multi-ingredient or specially engineered product rather than tested alone against a plain vehicle (PMID 39376788, PMID 30411631).

Is niacinamide good for acne?+

The evidence is thinner than the marketing. A 1995 double-blind trial randomised 76 patients, 38 to 4 percent nicotinamide gel and 38 to 1 percent clindamycin gel, for eight weeks. Improvement rates were 82 percent versus 68 percent on physician global evaluation, but the difference was not statistically significant (P equals 0.19) and there was no placebo arm, so it shows no detected difference between two active gels rather than proof either beat doing nothing (Shalita et al., PMID 7657446). The 2020 Cochrane review found four nicotinamide trials, none of which measured the review outcome for treatment response, which was the participants own global self-assessment (Liu et al., PMID 32356369).

Is niacinamide the same as niacin?+

No. They are two forms of vitamin B3. Niacinamide, also written nicotinamide, is the amide. Niacin, also written nicotinic acid, is the other form and the one the two are routinely confused for each other (Huber and Wong, 2020, PMID 33196157). The flushing reaction people associate with vitamin B3 supplements is attributed to nicotinic acid acting at the GPR109A receptor (Stern, 2007, PMID 21291680). On an ingredient list, niacinamide and nicotinamide mean the same molecule.

Can you use niacinamide with vitamin C?+

I searched PubMed for a controlled human trial testing whether layering the two cancels either one out and did not find one, so I cannot verify that claim from published human evidence. What does exist is a randomised, double-blind trial of a single product containing 10 percent nicotinamide alongside 5 percent magnesium ascorbyl phosphate, a vitamin C derivative, which was reported as safe and well tolerated over 60 days (Barbosa et al., 2024, PMID 39376788). That is not a direct test of layering two separate products, and one formulation does not settle the question for all of them.

Does taking niacinamide as a supplement do the same thing as a serum?+

No, and the two bodies of evidence should not be mixed. The oral trials tested 500 mg of nicotinamide twice daily for 12 months in people who had already had at least two skin cancers in the previous five years, under dermatologist follow-up. In immunocompetent high-risk patients that reduced new non-melanoma skin cancers by 23 percent (ONTRAC, PMID 26488693), while the same regimen in organ transplant recipients showed no benefit (ONTRANS, PMID 36856616). None of that describes what a topical serum does to the appearance of skin, and whether to take an oral form is a decision for a doctor, not a product page.

Research And Sources

Every figure on this page comes from one of the records below, read as published abstracts through the NCBI E-utilities interface. Where an abstract does not report a number, this page says so rather than filling the gap.

Nicotinamide: An Update and Review of Safety & Differences from Niacin · Huber R, Wong A · Skin Therapy Letter (2020) PMID:33196157
The role of nicotinic acid metabolites in flushing and hepatotoxicity · Stern RH · Journal of Clinical Lipidology (2007) PMID:21291680
Niacinamide: mechanisms of action and its topical use in dermatology · Wohlrab J, Kreft D · Skin Pharmacology and Physiology (2014) PMID:24993939
Nicotinamide increases biosynthesis of ceramides as well as other stratum corneum lipids to improve the epidermal permeability barrier · Tanno O, Ota Y, Kitamura N, Katsube T, Inoue S · British Journal of Dermatology (2000) PMID:10971324
The effect of niacinamide on reducing cutaneous pigmentation and suppression of melanosome transfer · Hakozaki T, Minwalla L, Zhuang J, et al. · British Journal of Dermatology (2002) PMID:12100180
The effect of 2% niacinamide on facial sebum production · Draelos ZD, Matsubara A, Smiles K · Journal of Cosmetic and Laser Therapy (2006) PMID:16766489
Topical niacinamide reduces yellowing, wrinkling, red blotchiness, and hyperpigmented spots in aging facial skin · Bissett DL, Miyamoto K, Sun P, Li J, Berge CA · International Journal of Cosmetic Science (2004) PMID:18492135
Niacinamide: A B vitamin that improves aging facial skin appearance · Bissett DL, Oblong JE, Berge CA · Dermatologic Surgery (2005) PMID:16029679
Niacinamide-containing facial moisturizer improves skin barrier and benefits subjects with rosacea · Draelos ZD, Ertel K, Berge C · Cutis (2005) PMID:16209160
Topical nicotinamide compared with clindamycin gel in the treatment of inflammatory acne vulgaris · Shalita AR, Smith JG, Parish LC, Sofman MS, Chalker DK · International Journal of Dermatology (1995) PMID:7657446
Topical azelaic acid, salicylic acid, nicotinamide, sulphur, zinc and fruit acid (alpha-hydroxy acid) for acne · Liu H, Yu H, Xia J, et al. · Cochrane Database of Systematic Reviews (2020) PMID:32356369
A Phase 3 Randomized Trial of Nicotinamide for Skin-Cancer Chemoprevention (ONTRAC) · Chen AC, Martin AJ, Choy B, et al. · New England Journal of Medicine (2015) PMID:26488693
Nicotinamide for Skin-Cancer Chemoprevention in Transplant Recipients (ONTRANS) · Allen NC, Martin AJ, Snaidr VA, et al. · New England Journal of Medicine (2023) PMID:36856616
Nicotinamide for Skin Cancer Chemoprevention (retrospective US Veterans Affairs cohort, with published correction) · Breglio KF, Knox KM, Hwang J, et al. · JAMA Dermatology (2025) PMID:40960808
Efficacy and Safety of Nicotinamide 10%, Associated with Magnesium Ascorbyl Phosphate 5% and Hyaluronic Acid 5%, Compared to Hydroquinone 4% in Women with Facial Melasma · Barbosa M, de Amorim RP, Cassiano D, et al. · Clinical, Cosmetic and Investigational Dermatology (2024) PMID:39376788
Novel nicotinamide skin-adhesive hot melt extrudates for treatment of acne · Nasr M, Karandikar H, Abdel-Aziz RTA, Moftah N, Paradkar A · Expert Opinion on Drug Delivery (2018) PMID:30411631