The White Cast Is the Protection. They Are the Same Thing.

Twenty-five percent of skin-of-color respondents in a sunscreen survey named white cast as a concern. Seven percent of non-Hispanic white respondents did.

That gap is the cleanest number available on the thing this whole subject is about, and the physics underneath it is worth understanding, because it explains why the problem has not been solved and probably cannot be.

Why it goes gray

Titanium dioxide and zinc oxide are semiconductors with band gaps in the ultraviolet. Below the band-gap wavelength they absorb light. Above it — across all of visible light — they cannot absorb at all. They can only scatter.

Scattering is efficient when a particle’s diameter is a substantial fraction of the wavelength it is meeting. Visible light runs 400 to 700 nanometers. Particles of roughly 150 to 300 nanometers scatter it hard, which is the definition of a white pigment. Non-nano titanium dioxide in sunscreen sits at 150 to 300 nanometers; zinc oxide at 200 to 400.

So the cast is not a formulation failure. The cast is visible-light protection, observed. A 2026 review states it about as plainly as it can be stated: for a sunscreen to protect against visible light, it should be perceptible on the skin — tinted, or producing a white cast — representing the reflection of visible photons to the retina.

An earlier review from the Henry Ford group puts it in one sentence: for a sunscreen to protect against visible light, it must be visible on the skin.

Now the trade. Mill the particles below about 100 nanometers and visible-light scattering collapses — that is why nano mineral sunscreens are clear — while the sub-band-gap ultraviolet absorption survives. You get a transparent product that still handles UVB and shorter UVA, and has effectively stopped doing the visible-light job.

The same 2026 review says it directly: formulations have overcome the cosmetic limitation by using nanoparticulate zinc oxide and titanium dioxide, and these are transparent and do not adequately protect against visible light.

There is a related caution about the word nano on a label. One analysis of a commercial 30-nanometer zinc oxide found that the primary crystallites were indeed 30 nanometers, and that in the actual dispersion they existed as aggregates spanning 0.6 to 60 micrometers. Nano describes the crystallite, not the optical unit the skin sees.

Why visible light matters here specifically

Mahmoud and colleagues, 2010, exposed twenty volunteers to either long-wavelength UVA or visible light on the lower back and followed the skin over two weeks with visual examination, cross-polarized photography, diffuse reflectance spectroscopy, confocal microscopy and biopsy.

In skin types IV to VI, both wavelengths induced pigmentation, and the pigmentation from visible light was darker and more sustained.

In skin type II, no pigmentation was observed at all.

That is the entire argument in one experiment. An untinted sunscreen is adequate for a type II and inadequate for a type V, and the reason is not preference or thoroughness. It is that visible light does something to melanocompetent skin that it does not do to lightly pigmented skin, and the standard broad-spectrum label stops at 400 nanometers.

The three percent rule, which does not exist

Iron oxides are what tinted sunscreens use to handle visible light — they absorb it, where pigmentary titanium dioxide scatters it, and the two work on the same problem from opposite directions.

Somewhere along the way a rule appeared: you need at least 3 percent iron oxide.

I went looking for the study that established it. There is no concentration series identifying 3 percent as a minimum effective dose. There is no threshold paper.

What there is, most likely, is a laundering. A 2025 randomized double-blinded trial had participants choose foundation shades matching their own skin tone, and as summarized in a later review this yielded iron oxide concentrations ranging from 3 to 27 percent. Three percent is the bottom of an observed shade-dependent range in one trial. It has become a formulation rule.

And the actual evidence on concentration disagrees with itself, three ways.

One 2025 study across 30 products and nine randomized trials reported better visible-light attenuation with higher iron oxide, in the range 6 to 25 percent.

A 2020 study in Fitzpatrick IV subjects reported similar protection across formulations with varying iron oxide concentrations — and, usefully, found the iron oxide formulations significantly outperformed both untreated skin and a non-tinted mineral SPF 50-plus.

And a 2024 study of twelve volunteers, irradiated at 320 joules per square centimeter across 370 to 700 nanometers and assessed at four timepoints by investigator grading, colorimetry and diffuse reflectance spectroscopy, found that the product giving the highest and most consistent protection contained 11 percent titanium dioxide and 1 percent iron oxide — beating a product with 12 percent zinc oxide, organic filters and 4 percent iron oxide.

A four-times-higher iron oxide load lost to a 1 percent load. Whatever the right answer is, a 3 percent minimum is not consistent with it.

The 2026 review concludes the obvious thing: further studies are needed to determine the minimum quantity of iron oxide necessary for effective visible-light protection. Its own limitations table says more studies are needed to determine the optimal concentration. And it flags the confound that makes all the comparisons soft — the formulations tested also contained varying titanium dioxide, which almost certainly contributed, so nothing here attributes the effect to iron oxide alone.

And you could not act on the rule anyway

A 2026 product analysis examined 37 tinted sunscreens, including contacting manufacturers directly.

Ninety-seven point three percent did not list iron oxide as an active ingredient.

Of the brands that responded, 9.6 percent disclosed a percentage when asked. The disclosed concentrations, where given at all, ran from under 1.4 percent to 10.4.

Six brands claimed visible-light protection. One of them publicly detailed any corresponding testing.

So even if a threshold existed, the number is not on the label, iron oxide is not regulated as an active ingredient in the US, and asking the manufacturer works about one time in ten.

There is a measured benchmark worth knowing, from a study using diffuse transmittance spectroscopy on formulations combining zinc oxide, titanium dioxide and iron oxide: 71.9 to 85.6 percent attenuation across 415 to 465 nanometers. That study’s formulations are one company’s products and the authors have industry ties, which belongs in the sentence. The review-level figure — that tinted sunscreens reduce high-energy visible transmission by up to 80 to 97 percent — carries an up to, and an up-to is a ceiling rather than a typical result.

The trial that matters for melasma

Sixty-eight patients randomized, 61 completing, eight weeks, intense sun in Mexico. Both arms got SPF 50 or above and both got 4 percent hydroquinone. The only difference was whether the sunscreen contained iron oxide to absorb visible light.

At eight weeks the visible-light arm had 15 percent greater improvement on the melasma severity index, 28 percent greater on colorimetry, and 4 percent greater on histological melanin.

Same drug, same SPF number, and blocking past 400 nanometers moved the outcome by a quarter on the instrument. It is the one intervention with a trial showing it improves the outcome of the melasma treatment you are also doing — which is why it sits at the floor of any melasma plan, under the topicals and well under the laser.

Note the 4 percent, though. The tissue-level difference was much smaller than the clinical and instrumental ones. That gap is worth sitting with rather than skipping.

What the dermatologists are doing

A survey of 974 practitioners found 91.68 percent actively counsel patients about visible-light protection.

Of those, 10.34 percent recommend visible-light-protective additives specifically for patients with melanin-rich skin. Forty-eight point eight nine percent recommend them for melasma or post-inflammatory hyperpigmentation.

So the advice is overwhelmingly attached to a diagnosis rather than to the skin type that the 2010 experiment says is at risk in the first place.

Eight point three two percent do not counsel on it at all, and the reason they give is the absence of standardized guidelines. The authors agree: there are currently no standardized recommendations for visible-light protection with these additives, leading to varied clinical recommendations.

That is the regulatory vacuum in one sentence. No threshold, no label requirement, no guideline, and an active ingredient that is not classified as one.

What changed in June 2026

The FDA added bemotrizinol to the sunscreen monograph at concentrations up to 6 percent — the first new US filter in roughly two decades. It is the ingredient sold elsewhere as Tinosorb S and available in Europe, Australia and Japan for years.

Here is the part worth noticing. It did not come through the Sunscreen Innovation Act, the 2014 law written specifically to unblock this queue. Every other filter sitting in that queue — bisoctrizole, drometrizole trisiloxane, enzacamene, amiloxate, octyl triazone, diethylhexyl butamido triazone — has had no Federal Register action since 2015. Eleven years.

Bemotrizinol advanced through the monograph order pathway created by the CARES Act in 2020, which was not written for sunscreen alone.

And the newer European filter designed to cover blue light, TriAsorB, shows its strongest genoprotective effect in combination with — bemotrizinol. The filter the US just approved is the one the next filter works best alongside.

What is still not measured

Two gaps, and they are both the same shape.

Nobody has published a measured UVA1 protection factor as a function of zinc oxide particle size in a finished product. The physics predicts that nano zinc oxide loses coverage in the 370 to 400 nanometer band, because zinc oxide’s absorption edge sits near 380 nanometers and above that it can only scatter — which small particles do badly. It is sound, it is widely accepted, and it has not been measured and published with a number. So nobody should print a percentage loss, including me.

And no study has systematically measured SPF or UVA protection factor performance against the skin tone of the test panel, nor validated any aesthetic acceptability instrument across Fitzpatrick I through VI. SPF testing panels are not required to be tone-diverse.

The product is tested for a number, the number is not tested across the people who will wear it, and the thing that determines whether anyone wears it — how it looks — has no instrument at all.

Choosing the cast

If your skin pigments in response to light, the sunscreen that works is one you can see. A tint that matches is the version of visible that is wearable; an untinted mineral that disappears is the version that has stopped working on the part you needed.

Do not chase a 3 percent iron oxide figure. It has no source, the evidence conflicts in both directions, and the number is not disclosed on 97 percent of products anyway.

The available shade ranges are the real constraint, and a content analysis of tinted sunscreen reviews found exactly that: most products offered only one shade, tone incompatibility was the most commonly cited negative, and the vast majority of those comments came from consumers with dark skin tones. The top-rated products were the ones with more shades. That is review-scraping rather than a controlled study, and it is also not surprising.

A cast you chose is a filter. A cast you did not is a product that was not made for you.

This article is educational and is not a substitute for care from someone who can look at your skin. Anything new, changing, painful or spreading should be assessed rather than researched.