Identity
Titanium dioxide (TiO₂) is an inorganic compound used as a UV filter — one of the two "mineral" or "inorganic" filters, the pair to zinc oxide. As a UV filter it is listed on labels as Titanium Dioxide; the same molecule, used as a white pigment, carries the Colour Index name CI 77891, which is why both appear on ingredient lists.
A note on how sunscreen is regulated, because it shapes what follows: in the US, sunscreens are over-the-counter drugs (they make a UV-protection claim); in the EU and UK they are cosmetics with an approved filter list in Annex VI. This is why the same ingredient carries a "GRASE" designation in one jurisdiction and an Annex listing in the other.
How it works
Titanium dioxide attenuates ultraviolet radiation through a combination of absorption, reflection and scattering — the same mechanism as zinc oxide, and the same practical consequence: the filter sits on the skin surface rather than being taken up by it.
Where it differs from zinc oxide is spectral reach. Titanium dioxide is strong in UVB and short-wave UVA (UVA-II) but tails off in the long UVA-I wavelengths — the deep-penetrating range most associated with photoageing and pigmentation — which zinc oxide covers better. So titanium dioxide is an excellent mineral UVB filter, but not a genuinely broad-spectrum one on its own; formulators pair it with zinc oxide or an organic UVA filter to close that gap. That single spectral shortfall is the reason this entry lands at B where zinc oxide lands at A.
The two honest caveats
Inhalation vs topical (IARC 2B). Titanium dioxide dust is classified by IARC as Group 2B — possibly carcinogenic to humans — on the basis of inhalation. This is important to state plainly and just as important to place correctly: it is a finding about breathing the airborne powder, in occupational settings, not about a cream applied to intact skin. The exposure route is different in kind. It is exactly why SCCS declines to recommend nano mineral filters in sprayable and powder formats — where an aerosol could be inhaled — while raising no such concern for creams and lotions. The honest framing is the same one the oxybenzone entry uses for absorption data: state the finding, and don't let a hazard from one exposure route be silently applied to another.
"Mineral means reef-safe" doesn't hold — but the coral data actually favour titanium dioxide here. Controlled work found zinc oxide nanoparticles caused coral bleaching, while titanium dioxide nanoparticles did not. That does not make titanium dioxide "reef-safe" — a marketing category, not a scientific one — but it is a genuine, specific point of difference worth recording rather than flattening into a slogan. And the attribution point stands: dermatology reviews are explicit that climate change is the dominant driver of coral decline, with sunscreen a local-scale contributor at most.
Reading the evidence honestly
What is genuinely strong. Excellent UVB and UVA-II protection, photostability, negligible skin penetration, and the best regulatory standing available — GRASE Category I in the US at up to 25% and the same ceiling in the EU. As a mineral UVB filter, titanium dioxide is first-rate.
Why it is B and not A. Only the long UVA-I shortfall. Zinc oxide is the one single filter that covers the entire UV range alone; titanium dioxide does not, so a titanium-dioxide-only product needs a UVA partner to be truly broad-spectrum. The grade reflects that gap, nothing more — it is not a safety demotion.
What the marketing gets wrong is the same thing it gets wrong for zinc oxide: that mineral filters are safe and chemical filters are not. Category III means missing data, not demonstrated harm. Titanium dioxide's standing is strong on its own terms; it is not evidence against the organic filters.
And the constant that outweighs filter choice. Neither spectrum nor filter class determines real-world protection as much as quantity applied and reapplication frequency. Most people apply well under the amount used to set a product's SPF, which makes the labelled number closer to a ceiling than a description.