Eye Care — Regulatory & Compliance Guide

Key Technical Parameters #

Getting eye area products to market is slower than most brand owners expect — and the bottleneck is almost never formulation. It’s documentation. The periorbital zone sits at the intersection of cosmetic and quasi-drug classification in several markets, which means the compliance burden shifts depending on where you sell, what claims you make, and what actives you use. Brand owners who develop clean files from day one move through registration 30–40% faster than those who retrofit documentation after the fact. This guide focuses on the pre-market documentation architecture — what you need, in what sequence, and where the gaps typically appear.

What the Markets Actually Require: EU, US, and China Side-by-Side #

The regulatory floor for eye area cosmetics differs more between markets than most people realize. Not in a vague “different standards” way, but in specific, document-level ways that affect what we build into the product file from batch one.

The EU Cosmetics Regulation 1223/2009 requires a Product Information File (PIF) before any product is placed on market. For eye area products specifically, the Safety Assessment under Annex I must address periorbital sensitization risk explicitly — not just the general NOAEL/MoS calculation. We’ve had assessors come back on eye cream files that had perfectly adequate general safety data but lacked a specific commentary on the ocular mucosa exposure route. That revision cost one client six weeks.

The US sits on the other end. Under current FDA Cosmetics Guidelines, eye area products are cosmetics unless they make drug claims — no pre-market notification, no mandatory safety assessment on file with the agency. What the FDA does require is that your product be safe for its intended use and that labeling comply with 21 CFR Part 701. The practical implication: your safety burden is real, but it sits with you, not with a regulator reviewing a submission.

China is the most demanding pathway. NMPA requires full registration for imported general cosmetics, including eye products. The NMPA Cosmetic Regulation framework mandates human safety testing conducted at a Chinese NMPA-accredited lab, a certificate of free sale from the country of origin, and — if you’re using any new cosmetic ingredients — a separate ingredient notification process that can add 6 to 12 months to the timeline.

That last row is where projects stall. A brand that wants simultaneous EU and China launch with a novel active — say, a synthetic peptide at 5 ppm — needs to account for up to 18 months of China lead time before any commercial inventory ships. We flag this in every kickoff call. Launches get planned around EU timelines and then the China channel gets delayed by a year.

Where the Documentation Falls Apart: Root Causes We See Repeatedly #

This is where most of the real project pain lives.

The ophthalmologist test gap. EU law doesn’t mandate an ophthalmologist test for eye cream, but in practice, any competent EU safety assessor will ask for one if the product makes proximity claims to the eye or if the formula contains anything with a potential ocular irritation profile. We use the Draize ocular scoring protocol as a reference framework, though the actual test is conducted on human subjects in a clinical setting — typically 30 subjects, single-application observation at 15 minutes, 1 hour, and 24 hours. A product with a score above 1 on the modified Draize scale at the 24-hour read will not pass a credible EU safety assessment. We’ve seen this happen with undiluted fragrance compounds carried over from face cream versions of the same formula.

The mistake is usually scope compression. A brand ports a face cream formula into an eye cream SKU, changes the name, and assumes the existing safety data covers it. It doesn’t. The exposure geometry around the periorbital area — proximity to conjunctiva, thinner stratum corneum, higher absorption — means the MoS calculations need to be recalculated from scratch using eye-specific exposure assumptions (typically 0.2–0.5 g per application versus 1.0–2.0 g for a face product). That difference in applied weight alone changes the MoS by a factor of 2 to 4x on a worst-case basis.

The INCI declaration mismatch. China’s IECIC (International Cosmetic Ingredient Codex) is a positive list — if an ingredient isn’t on it, it cannot be used without a new ingredient notification. We routinely audit incoming briefs for IECIC coverage before we start formulation, using what we call our CIL-3 ingredient pre-screen. In 2024, across 31 eye area briefs we received from overseas brands, 14 contained at least one ingredient not on the current IECIC list. Usually it’s a trademarked peptide or a novel ferment extract. The brand had already featured it in marketing materials. Reformulation at that stage is expensive and annoying.

Claim-triggered reclassification. This one surprises people. An eye product that claims to “reduce under-eye puffiness” is a cosmetic claim in the EU and the US. An eye product that claims to “improve lymphatic drainage” or “treat periorbital edema” can be interpreted as a medical device or drug claim, depending on the market and the specific wording. In France specifically, we’ve seen products held at customs because the claim language on the shipper’s invoice used the word “traitement” without qualification. The products were cosmetically formulated. The word choice was the problem. Under SCCS Scientific Opinion guidance, any claim that references a physiological mechanism at the tissue or cellular level gets scrutinized more closely for reclassification risk.

Stability documentation for actives at low pH. Some eye area actives — ascorbic acid derivatives, certain AHAs used at low concentrations — require stability data at accelerated conditions: 40°C / 75% RH for 12 weeks minimum, per ICH Stability Guidelines. For eye products specifically, we also run a freeze-thaw cycle (3 cycles, -15°C to 25°C) because distribution through cold chain environments is common in markets like South Korea and parts of Northern Europe. A product that looks stable at ambient conditions can show phase separation or active degradation after freeze-thaw. We’ve had batches of a peptide eye serum pass 12-week accelerated stability and then fail freeze-thaw at cycle 2. The culprit was a co-emulsifier that destabilized at sub-zero temperatures. That cost a client a full reformulation cycle and pushed their China registration filing back by four months.

Does “Ophthalmologist-Tested” Actually Mean Anything on Pack? #

Yes — but only if the test protocol is documented and the claim is scoped correctly.

“Ophthalmologist-tested” as a pack claim requires that an ophthalmologist was involved in designing or reviewing a clinical tolerance study. A single dermatologist sign-off on a HRIPT doesn’t satisfy this. The test typically needs to involve ocular tolerability assessment, not just skin sensitization. The PCPC Guidelines provide useful framing for what “tested” implies from a claim substantiation standpoint, though they’re non-binding. In the EU, if the claim appears on pack, it needs to be substantiated in the PIF under Article 20 CPR. A 2022 controlled tolerance study (n=52, 4 weeks, twice-daily application at the periorbital site) reported 98% subject tolerability with zero adverse ocular events — that’s the kind of documentation that makes an Article 20 file defensible. For any eye area product making this claim, we require the study report to be in our possession before we issue a Certificate of Conformity.

That said, the claim does carry weight with consumers. Whether it actually predicts product safety better than a well-constructed safety assessment is a different question. Our view: the ophthalmologist test is more valuable as documentation infrastructure than as a performance differentiator.

Formulation Notes for Brand Partners #

When you brief us on an eye area product, the first questions we ask are: What markets are you registering in — and in what sequence? What is your on-pack claim language, and has legal reviewed it? Those two questions determine more about the documentation workload than anything in the formula itself.

The most common brief mistake we see is treating eye area documentation as an afterthought — assuming you can build the compliance file after the formula is locked and the launch date is set. What actually happens is that the safety assessment comes back with a question about a specific ingredient at a specific concentration, and you’re reformulating at week 14 instead of week 4.

We also push back when brands submit briefs with claim language that was written by a marketing team without regulatory input. “Lifts and firms the eye contour” is fine. “Stimulates collagen synthesis in the periorbital dermis” is not — at least not without ISO Standards-aligned clinical data to back it. We’ll tell you this before we start, not after.

For eye care products targeting EU and China simultaneously, our standard timeline is: lab samples in 2–3 weeks, accelerated stability initiated at 4 weeks, ophthalmologist tolerance study commissioned at week 6 (running concurrent to stability), and China registration filing submitted at week 18 at the earliest. Twenty-four-month real-time stability runs from batch one. Build that timeline into your launch calendar before you commit to retail listings.

Frequently Asked Questions #

We want to launch in the EU first, then China. Can we use the same formula for both?
A: Usually yes on the formula, but not on the documentation. EU and China have different safety data format requirements, different ingredient list restrictions, and China requires human testing at an NMPA-accredited facility that your EU assessor’s data won’t substitute for. Budget for parallel documentation tracks from the start.

Our marketing team wrote the claims already. Do we need to change them before filing?
A: Bring them to us at brief stage. Claim language is the single fastest way to trigger a reclassification review in the EU or a rejection from NMPA. We’ve seen products delayed six months over a single phrase. Under EU Cosmetics Regulation 1223/2009 Article 20, every on-pack claim needs substantiation in the PIF — if the claim is ambitious, the evidence burden goes up accordingly.

What happens if our eye product fails the ophthalmologist test?
A: Then you reformulate before filing, not after. A Draize-equivalent ocular score above 1 at the 24-hour read is a hard stop for most EU safety assessors. The typical cause in our experience is fragrance or a sensitizing preservative carried over from a face cream version of the same product. Our encapsulation technology can reduce free-phase concentration of certain actives, which sometimes resolves borderline irritation results — but it depends on what’s causing the score.

What’s your MOQ and how long does sampling take for eye area products?
A: Sampling MOQ for pilot batches is typically 200–500 units depending on format. Full production MOQ ranges from 3,000 to 10,000 units depending on packaging complexity. Lab samples in 2–3 weeks from brief sign-off, with accelerated stability data available at 8 weeks.

Is there anything in our documentation file that brands consistently forget to include?
A: The Certificate of Free Sale — consistently. China NMPA requires one issued by the competent authority in the country of origin, and getting it takes 4–8 weeks depending on the issuing body. Brands brief us, we lock formula and file, and then the China submission stalls for two months waiting for a piece of paper that could have been requested on day one. Request it in parallel with formulation, not after.

---

Have a product concept in mind? Contact our formulation team to request a complimentary brief review.