Niacinamide is one of the best-studied ingredients in skincare, which makes the marketing around it especially odd: the percentages printed on bottles have drifted far away from the percentages used in the research.
What the studies actually used
The landmark trials behind niacinamide's reputation — on barrier function, hyperpigmentation, blotchiness, and fine lines — were run at 2% to 5%:
- 2% improved barrier function and reduced water loss in several studies, and reduced sebum production in a Japanese trial on oily skin.
- 4% reduced hyperpigmentation and blotchiness over eight weeks, and performed comparably to 4% hydroquinone in one melasma study with fewer side effects.
- 5% improved fine lines, elasticity and skin tone over twelve weeks.
There is no published evidence that 10% outperforms 5% for any of these outcomes. None. The 10% figure exists because it is a bigger number on a front label, and because one famous product made it the default.
Does 10% hurt?
For most people, no — niacinamide is well tolerated. But the side-effect rate is not flat with dose: flushing, redness and irritation reports climb at higher concentrations, and the conversion of niacinamide to nicotinic acid (the flushing pathway) is more likely in formulas with a low pH. If your skin stings from a 10% serum, dropping to a 4–5% formula usually keeps the benefit and loses the reaction.
There is also an opportunity cost. Every percentage point of niacinamide is a percentage point that is not another active. A 10% niacinamide serum is often 10% of an ingredient you only needed 4% of.
How to read a niacinamide serum
- 2–5% niacinamide with supporting ingredients (zinc PCA for oil control, tranexamic acid for pigmentation) is the evidence-aligned zone.
- Above 5% is not dangerous, just unproven — treat the extra as marketing unless a specific study says otherwise.
- Position on the INCI list matters when no percentage is disclosed: niacinamide before the preservative line (phenoxyethanol, usually capped at 1%) means a meaningful dose; after it usually means under 1%.
That last point is exactly the logic Skinparency's serum scoring uses: where a brand publishes a concentration we use it, and where it does not we estimate from the ingredient's position relative to known 1% markers — and label the result as an estimate rather than a fact. An underdosed serum is not a bad product, but you should not pay a premium for a name on the label.