Why La Roche-Posay Anthelios Scores So Low on Yuka — and Whether It Should Change What You Buy

Anthelios Melt-in Milk Sunscreen SPF 60 — the La Roche-Posay sunscreen that American dermatologists recommend more than almost any other — scores 0 out of 100 on Yuka. Not a low score. The floor.

That reading was taken in the Yuka app on 12 August 2026, on the US product. It is not an outlier, and it is not a European quirk of ingredient law: the day before, we read eight La Roche-Posay products on the French market and they came back between 1 and 51 out of 100, seven of them Anthelios sunscreens.

So the pattern holds across two completely different formulations, sold under two different regulatory regimes. That is a stronger finding than either half on its own, and it is worth understanding properly, because the honest answer to "should I stop using this?" is not the answer most people assume when they see a zero.

This guide does four things: it shows the scores with their provenance, explains the specific piece of arithmetic that produces a 0 on a well-regarded SPF 60, adjudicates whether Yuka is right, and tells you what we would buy.

The Scores, and Exactly Where They Came From

Every number below was read directly in the Yuka app on the date shown. We do not estimate scores, and we do not carry a score across from a similarly-named product — which matters enormously here, because La Roche-Posay sells different Anthelios formulas in different markets under confusingly similar names.

United States

France

These eight are reproduced with the product names exactly as the app displayed them, ellipses included. Five of the eight names were truncated by the app's own interface, and we have deliberately not guessed at what the missing characters were.

All eight read in the Yuka app on 11 August 2026. Seven of the eight are Anthelios sunscreens; the outlier is a face wash.

Two things we want to be upfront about, because they limit what these numbers can be used for.

• We could not confirm which exact SKU most of these rows refer to. We searched all 118 La Roche-Posay records in the OpenBeautyFacts database and could not confirm a one-to-one match for a single one of the eight French products. Only the Dermo-Pediatrics spray has a probable match, and "probable" is not "confirmed". So the scores are published against the names Yuka displayed, and nothing more is claimed.
• The two products scoring 23 are not the same product, and we are not going to pretend to know which is which. Both are sprays with truncated names.

Note also what the bands do. 50 and 51 render green, labelled "Good", while 43 renders as "Poor". Yuka's Good boundary sits at 50, so a one-point difference flips the colour of the badge. Read the number, not the colour.

Why a Good Sunscreen Can Score 0: The Ceiling

This is the part that explains the whole table, and almost nobody who sees a score in the app knows it. Yuka's cosmetics score is not an average of how good the ingredients are. It is a ceiling set by the single worst ingredient in the formula, and then penalties applied underneath that ceiling.

From Yuka's own published methodology (help.yuka.io, "How are penalties calculated in the cosmetic product scores?", which we re-read on 12 August 2026):

• A product containing a high-risk (red) ingredient scores "between 0 and 24/100".
• A product containing a moderate-risk (orange) ingredient — and no red one — scores "between 0 and 49/100".
• A product containing only low-risk and no-risk ingredients: "The product's score will systematically be greater than or equal to 50/100".

Penalties are then subtracted within that band. For a formula that already contains an orange or red ingredient, they run −12 for a red carcinogen or endocrine disruptor, −8 for a red allergen or irritant, −6 and −4 for the orange equivalents, −3 and −2 for yellow. Where an ingredient carries several risks, only the highest applies. Products with three ingredients or fewer are penalised on an adjusted basis, because each ingredient is a larger share of the whole.

Three consequences follow, and they change how the whole table should be read.

1. Everything else in the bottle is irrelevant to the ceiling. A sunscreen can be photostable, broad-spectrum, fragrance-free, non-comedogenic and clinically tested on reactive skin, and none of that lifts it above 24 if one ingredient is red.
2. The gap between 0 and 23 is not a safety gradient. Both sit under the same ceiling. The difference between them is how many additional flagged ingredients happen to be present — not how dangerous the product is.
3. A score of 50 or 51 does not mean "moderately safe". It means no ingredient was flagged orange or red at all, and then a couple of points came off. Those two products are on the other side of a categorical line, not slightly further along a continuous one.

This is why the headline number is close to useless on its own, and why the interesting question is never "what is the score" but "what got flagged, and was flagging it correct?"

What Is Actually in the Bottle

Yuka does not publish which specific ingredient it flagged in a given product, or how many points each one cost. We are not going to guess, and anyone claiming to have reverse-engineered it from a handful of products is guessing. What we can do is state what is in these formulas, from primary sources, and then say which of those ingredients carries a documented regulatory concern.

The US formula

La Roche-Posay's own US site lists the actives in Anthelios Melt-in Milk SPF 60 as avobenzone 3%, homosalate 10%, octisalate 5% and octocrylene 7%. That is a textbook FDA-monograph filter set: four organic filters, no mineral filter at all.

The European formulas

The European Anthelios range is built differently. Pulling published ingredient lists for Anthelios products from the OpenBeautyFacts database on 12 August 2026 — for example the Dermo-Pediatrics SPF 50+ Spray Multi-Positions (barcode 3337872413407) and Anthelios XL 50+ ultra-léger (3337872414527) — the filter systems combine some of the same molecules with several that are unavailable in the United States:

• Shared with the US formula: homosalate, ethylhexyl salicylate (octisalate), octocrylene, butyl methoxydibenzoylmethane (avobenzone).
• European-only, or until very recently so: bis-ethylhexyloxyphenol methoxyphenyl triazine (bemotrizinol, sold as Tinosorb S), ethylhexyl triazone, drometrizole trisiloxane (Mexoryl XL) and terephthalylidene dicamphor sulfonic acid (Mexoryl SX).

The ingredient with a documented file against it

One filter appears in both the US and the European formulas and carries an actual regulatory history: homosalate.

The EU's Scientific Committee on Consumer Safety reviewed it over concerns about potential endocrine-disrupting properties. Its final opinion concluded that homosalate is safe as a UV filter only up to 0.5% of a finished sunscreen product. A follow-up scientific advice (SCCS/1638/21, adopted 2 December 2021) then addressed a narrower question — is it safe up to 7.34% in face products only, face cream and pump spray — and answered yes. The reasoning is exposure: a face-only product covers a fraction of the body, so the same percentage delivers far less substance into the body.

The EU legislated on that basis. Since 1 July 2025, homosalate may be sold in the EU only in face products, and only up to 7.34%.

Now put the American bottle next to that. The FDA's sunscreen monograph permits homosalate up to 15% (21 CFR 352.10), and Anthelios Melt-in Milk SPF 60 contains 10% — in a lotion marketed for face and body. That exact product, at that concentration and that use, could not legally be sold in the European Union today.

To be precise about what this does and does not establish: it does not prove that homosalate is the ingredient Yuka flagged. It does establish that the concern the app is encoding is not something the app invented.

Is Yuka Right? Our Verdict

This is the question worth having an opinion about, so here is ours, with the case against the app first.

Where the critics are correct

Cosmetic chemists have a consistent and well-founded objection, set out clearly in Hypebae's November 2025 interview feature on the app. Biochemist Mollie Kelly Tufman: Yuka's "main failure is that it leads users to oversimplify the toxicology process. It's not just about whether an ingredient is present — but how much, how it's used, and on what kind of skin." Formulator Dr Shuting Hu puts the mechanism plainly: "Two products can contain the same flagged ingredient, yet one can be perfectly appropriate because the dose, vehicle and use pattern keep exposure well below accepted safety margins." And Smitha Rao notes the structural problem underneath both — ingredient concentrations are generally not disclosed in cosmetics, so an app reading an ingredient list cannot know the dose in the first place.

That criticism lands hard here. The SCCS reached two different conclusions about the same molecule — 0.5% for a whole-body sunscreen, 7.34% for a face product — purely on exposure. A hazard-based flag cannot make that distinction, and the ceiling mechanism means it does not even try.

There is a second objection which is arguably more serious than any of this. A 0/100 on a sunscreen invites a reader to stop wearing sunscreen. Whatever the risk from a UV filter, it is not symmetrical with the risk from ultraviolet radiation, which is a known human carcinogen with an unambiguous dose-response relationship. An app that pushes a person from an imperfect sunscreen to no sunscreen has made them less safe, not more.

Where Yuka is on firmer ground than its critics allow

The usual rebuttal to a scary score is "regulators have assessed this and are satisfied." In the United States, on these specific ingredients, that rebuttal is not available.

In its sunscreen review, the FDA proposed that only two active ingredients are generally recognised as safe and effective: zinc oxide and titanium dioxide. Two more — PABA and trolamine salicylate — were proposed as not GRASE on safety grounds. And twelve others were proposed as not GRASE because additional safety data are needed, a list which includes homosalate, octisalate, octocrylene, oxybenzone and avobenzone: every single active in the Melt-in Milk bottle. Those ingredients remain legal to sell, and the agency has been asking for the missing data for years.

So the position is this. The EU restricted homosalate after a scientific review. The FDA has not certified any of these four filters as safe and effective and is still waiting on data. Yuka is pointing at a real, regulator-acknowledged gap. It is doing so with an instrument that is far cruder than the question deserves.

The verdict

Yuka is directionally defensible and quantitatively meaningless here.

It is right that these formulas rest on filters with an open safety file, and right that a shopper might reasonably want to know. It is wrong to express that as a number on a 0–100 scale, because the number is not a measurement of anything. A 0 and a 23 and a 34 differ by how many flagged ingredients are present, not by how risky the products are, and none of the three is a statement about your actual exposure. Treat the score as a flag, and specifically as a flag about UV filter chemistry rather than about La Roche-Posay's competence as a formulator.

And the practical corollary matters more than the verdict: the correct response to a low Yuka score on a sunscreen is to change sunscreen, never to skip it.

Why the US Bottle Is Genuinely a Different Product

If you are reading this in the United States, the European scores above are context, not information about your bottle. The two product families share a name and very little else, because the two regimes approve different filters.

The EU regulates sunscreens as cosmetics, with permitted UV filters listed in Annex VI of Regulation (EC) 1223/2009. The US regulates them as over-the-counter drugs under a monograph, which is a far slower instrument to update — and the difference is not academic:

• Bemotrizinol (Tinosorb S) has been permitted in the EU for years at up to 10%. The FDA added it to the OTC monograph at up to 6% only in June 2026 — "the first new active ingredient included in the OTC sunscreen monograph since the late 1990s", as COSlaw reported on 23 June 2026. It appears in the European Anthelios ingredient lists above; it is not among the actives in the US Melt-in Milk formula.
• Mexoryl SX and Mexoryl XL, L'Oréal's own filters and a large part of why European Anthelios has the UVA reputation it does, are not on the US monograph list at all.
• Conversely, homosalate at 10% in a body lotion — the US formula — has been off-limits in the EU since July 2025.

The practical rule: a Yuka score read on one market's SKU tells you nothing reliable about the other market's SKU. That is precisely why we read the US product separately rather than assuming the French numbers transferred, and why every row in the tables above names its market.

What We Would Actually Buy

This is the part with a decision in it. Four options, in the order we would consider them.

1. Switch to a mineral sunscreen — the only move that actually addresses the flag

If the open safety file on organic UV filters is what bothers you, the answer is a sunscreen whose actives are zinc oxide, titanium dioxide, or both. Not because mineral filters are "natural", but for one specific and checkable reason: they are the only two the FDA has proposed as generally recognised as safe and effective. Everything in this article's adjudication section stops applying to them.

The trade-off is real and you should know it before you switch: mineral formulas can leave a white cast, particularly on deeper skin tones, and they are generally less pleasant to reapply. A sunscreen you dislike is a sunscreen you underapply, and underapplied SPF 50 does not deliver SPF 50.

2. Use a face sunscreen on your face — a smaller, cheaper fix

The regulatory concern about homosalate is a whole-body exposure argument, which is exactly why the EU now permits it in face products and nowhere else. If you apply a large amount of sunscreen over a large area every day, that is where the exposure argument bites. If your sun protection is a face product used on a face, you are already inside the case the SCCS looked at and accepted at up to 7.34%.

3. Keep the brand for everything that is not sun care

Notice what the French table actually shows: seven of the eight low scores are sunscreens. The low scores are a category finding, not a brand finding. La Roche-Posay's non-sun products are a different proposition — our review of the Toleriane Double Repair moisturiser and of the Lipikar Cleansing Oil AP+ both look considerably better on ingredient grounds, and the full picture sits on the La Roche-Posay brand page.

4. Change nothing, deliberately

This is a legitimate option and it deserves stating alongside the others. Every product in the tables above is legally sold, dermatologist-recommended in its market, and demonstrably effective at the thing sunscreen is for. If you have a formula that suits your skin, that you will actually reapply, and you are unpersuaded by a hazard flag that cannot see dose — keeping it is a defensible choice. As Tufman put it in the same Hypebae feature: "The best product is one that's well-formulated, stable and suits your skin, not necessarily the one that gets a perfect rating on an app."

The Bottom Line

La Roche-Posay's sunscreens score badly on Yuka in both of the markets we have read — 0/100 for the US Melt-in Milk SPF 60, 1 to 51 across eight French products — and the mechanism is not mysterious. Yuka caps a product at the risk level of its worst ingredient, sunscreens are defined by their UV filters, and several of the most widely used organic filters have an open regulatory file on both sides of the Atlantic.

What the score cannot tell you is whether that matters at the dose in your bottle, on the area of skin you apply it to. It is a flag, not a measurement, and the gap between a 0 and a 34 carries almost no information.

If the flag bothers you, move to zinc oxide or titanium dioxide and accept the cosmetic trade-off. If it does not, keep what you have. The one response that is clearly wrong is the one a zero most tempts people towards, which is wearing nothing at all.

For contrast, the same method applied to a brand at the opposite end of the scale is worth reading: every Garnier product we have scored sits in Yuka's top band, and the reason is the same ceiling rule working in the other direction.

Sources

• Yuka scores: read in the Yuka app, US product 12 August 2026, French products 11 August 2026.
• Yuka scoring method: help.yuka.io, "How are penalties calculated in the cosmetic product scores?", re-read 12 August 2026.
• US actives: La Roche-Posay US product page for Anthelios Melt-in Milk Sunscreen SPF 60, retrieved 12 August 2026.
• European ingredient lists: OpenBeautyFacts, retrieved 12 August 2026 (barcodes cited in text).
• Homosalate: SCCS opinion on homosalate and SCCS/1638/21 scientific advice, adopted 2 December 2021; EU restriction in force 1 July 2025. US limit: 21 CFR 352.10.
• GRASE status of US sunscreen actives: FDA questions and answers on the sunscreen deemed final order and proposed order.
• Bemotrizinol: COSlaw.eu, "FDA updates sunscreen monograph with new UV filter", published 23 June 2026, re-read 12 August 2026.
• Expert criticism: Hypebae, "Cosmetic Chemists Break Down the Yuka App Trend", by Kleigh Balugo, 6 November 2025.