What comedogenic ratings actually measure

Somewhere on the internet is a table giving every cosmetic ingredient a number from 0 to 5. Coconut oil is a 4. Dimethicone is a 1. Shea butter is a 0 or a 2 depending on which table you found.

Those numbers are quoted more confidently than almost anything else in skincare, they appear in product reviews and shopping guides, and very few of the people using them know where they came from or what happened to them since.

Where the numbers came from

Rabbit ears.

The comedogenicity scale was built on a bioassay in which a substance was applied to the inside of a rabbit’s ear, which forms comedones readily, and the result was scored. It was a reasonable screening tool when it was developed in the 1970s and it produced the tables still in circulation today.

A 2026 dermatology review states the problem directly: “Historically, comedogenicity ratings were based on rabbit ear bioassays, which overestimate real-world likelihood of poor blockage in humans.”

Overestimate. The rabbit ear is more reactive than human facial skin, so a substance that produces comedones there may produce none on a person. The tables are not random, and they are systematically alarmist.

What replaced them

The same review is unambiguous about current practice: “Modern dermatology no longer relies on ingredient-level comedogenicity lists; current evidence supports evaluating whole formulations rather than single ingredients.”

That sentence contains the whole argument of this guide, and it has two halves worth separating.

Ingredient-level lists are the wrong unit. An ingredient at 20 percent in a heavy balm and the same ingredient at 0.5 percent in a light gel behave differently. Concentration, what it is combined with, and how the product is formulated all change the outcome. A single number attached to a molecule cannot carry any of that.

Formulations are the right unit, and you cannot look one up. There is no table of finished products, because a formulation has to be tested as itself and most are not tested publicly at all.

So the honest position is uncomfortable: the number you can find is not reliable, and the assessment that would be reliable does not exist for the product in your hand.

Non-comedogenic is not a regulated term

Worth stating plainly. In the United States there is no legal definition of non-comedogenic, no required test, and no threshold a product must meet before printing it.

It sits with “hypoallergenic”, “dermatologist tested” and “clean” in the category of terms that sound like standards and are not, which is set out in the guide to what natural and clean legally mean.

A product carrying the claim may have been tested carefully. Nothing on the label distinguishes that one from a product where marketing chose the phrase.

What is actually true about cosmetics and acne

This is where the guide has to be careful, because discrediting the ratings is not the same as saying cosmetics do not cause acne. They do, and it has a name.

A 2025 case-control study of 151 participants found that “facial cleansers ( P =0.04), foundation ( P= 0.03), and powders ( P= 0.01) were related with an increased risk of acne”. Acne cosmetica is real and was described half a century ago.

Notice what that study found: product categories, not ingredients. Foundation and powder are risk factors as products, which is exactly the formulation-level finding the review recommends and the ingredient tables cannot produce.

What to do instead

Judge the product, not the list. A heavy occlusive balm and a fluid gel differ in ways no ingredient table captures.

Use your own history as data. If a product broke you out, that product is the finding, and it beats any rating for you specifically. If one did not, its ingredients being on a list is not a reason to stop.

Be sceptical of the number in both directions. A high rating is not proof of trouble and a 0 is not a guarantee.

Change one thing at a time. Acne responds slowly and a routine changed wholesale tells you nothing about which part mattered.

The ingredients most unfairly treated

Dimethicone sits near the top of most avoid-lists and is inert, non-occlusive in the relevant sense, and among the least reactive things in cosmetics.

Petrolatum is heavily occlusive, which is its purpose, and its comedogenic reputation comes largely from a rating rather than from clinical evidence in humans.

Lanolin has a genuine contact allergy record, which is a different objection and a better documented one. Where an ingredient has real evidence against it, this site says so on its own page rather than through a number.