Eye irritation – OECD 492

OECD guideline 492 describes an in vitro test method performed on reconstructed human cornea-like epithelium (RhCE ). It allows for the determination of whether a substance or mixture can be excluded from any classification for eye irritation or serious eye damage, without resorting to animal testing.

The principle is based on a three-dimensional model of human corneal epithelial cells, cultured at the air-liquid interface until a stratified tissue forms that faithfully reproduces the corneal barrier. The product to be tested is applied to the surface of the tissue, and then cell viability is measured using the MTT assay. High viability after exposure indicates that the product has not significantly altered the corneal tissue.

Adopted in 2015, guideline 492 has undergone several revisions, the most recent dating from June 2025.It now constitutes the in vitro for assessing ocular risk.

What Article 492 allows — and does not allow — to conclude

This is the most frequently misunderstood point, and it conditions your entire testing strategy.

OECD 492 is an exclusion method, used in a bottom-up. It answers a single question, with a yes or no answer: can this product be classified as "unclassified"?

• Viability exceeding the model threshold → the product is identified as not requiring classification (No Category as defined by the GHS). The conclusion is final; no further testing is required.
• Viability less than or equal to the threshold → the product cannot be excluded from classification. Note: this does not it is classified as an irritant. Guideline 492 does not distinguish between category 2 (eye irritation) and category 1 (serious eye damage). Further testing is therefore required.

In other words: a negative result closes the case, a positive result opens it. It is precisely this limitation that OECD guideline 492B, as it distinguishes between the three categories in a single test.

Test procedure

The exact parameters—immersion and post-exposure incubation times, viability threshold— depend on the selected RhCE model. Each validated model has its own protocol and decision threshold, defined in the guidelines. Our partner laboratories will guide you toward the model best suited to the nature of your product.

Validated RhCE models

Several models of reconstituted corneal epithelium are covered by guideline 492, including the most commonly used in Europe:

• EpiOcular™ EIT (MatTek) — the historical model, the most widely documented;
• SkinEthic™ HCE EIT (Episkin) — a model of European origin, widely used in cosmetics.

Other models have been added through successive revisions. The choice of model is not neutral: it determines the applicable viability threshold, the exposure duration, and compatibility with certain matrices. We will verify this point with you before launching the study.

Which products and sectors are affected?

OECD test 492 applies to substances and mixturesin liquid, solid, semi-solid, or waxy form. Depending on the case, the following is required:

• Cosmetics and cosmetic ingredients — within the framework of Regulation (EC) No 1223/2009, which prohibits animal testing: the in vitro is the only possible one;
• Chemical substances — under REACH registration and CLP classification;
• Detergents and cleaning products — Regulation (EC) No 648/2004 and hazard labelling;
• Biocides — authorisation dossiers under the BPR regulation;
• Plant protection products, industrial raw materials, devices and materials that may come into contact with the eye.

Regulatory framework and CLP classification

The test result directly informs the classification under the CLP regulation, the European transposition of the GHS:

• No classification — the product is not subject to any eye hazard warnings;
• Eye Irrit. 2 (H319, "Causes serious eye irritation");
• Eye Dam. 1 (H318, "Causes serious eye damage").

OECD 492 only allows us to conclude in the first case. To decide between the other two, we must either follow up with a complementary method (OECD 437 BCOP, OECD 438 ICE), or adopt OECD492B.

OECD 492 or OECD 492B: which one to choose?

Our advice: if you have good reason to believe your product is non-irritating, OECD 492 is the fastest and most economical option. If in doubt, OECD 492B avoids a second test and saves several weeks. We explain this choice in detail in our article OECD 492 or 492B: which method to choose?

Why conduct this trial with YesWeLab?

• A network of laboratories specializing in in vitro, selected based on their real-world experience with RhCE models and their accreditations;
• The right model the first time — we check the compatibility of your matrix with the intended model before launching the study, which avoids inconclusive trials;
• A centralized quote request : a single point of contact, even when your case combines several guidelines;
• A study report that can be directly used in your REACH, CLP, cosmetics or biocide files.

Frequently Asked Questions

What is the difference between OECD 492 and OECD 492B?

Both methods use a reconstituted human corneal epithelium, but they do not address the same question. OECD 492 only identifies products that do not require classification. OECD 492B, through a multi-exposure approach, distinguishes between categories 1, 2, and no classification in a single test.

Is OECD 492 acceptable for a cosmetic case?

Yes. Regulation (EC) No 1223/2009 prohibits animal testing; in vitro such as OECD 492 are the normal way to assess ocular risk for cosmetic ingredients and finished products.

Does a positive result mean that my product is irritating?

No. A viability below the threshold simply means that the product cannot be excluded from classification. It may fall under category 2 or category 1—or, in some cases, not be irritating at all. Further testing is required to determine this.

How long does the test take?

The experimental phase takes several days. The overall timeframe, from sample receipt to report delivery, depends on fabric availability and the test plan. We will provide you with a firm timeframe along with the quote.

How much product do I need to supply?

A few grams or milliliters are sufficient in the vast majority of cases. The exact quantity depends on the number of replicates and the possible need for preliminary testing; it is specified in the quote.

My product is colored and spontaneously reduces MTT. Is this a problem?

This is a known interference with the method. It can be managed through additional controls outlined in the guideline, provided it is anticipated. Please inform us of any significant coloration or reducing effect when requesting a quote.

Request your quote

Describe your product and your regulatory objective: we identify the appropriate laboratory and model, and get back to you with a costed proposal.