TL;DR: In the EU, the same phrase sits in a grey zone under [EU Cosmetics Regulation 1223/2009](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32009R1223), where it’s permissible as a cosmetic claim but will trigger tighter scrutiny at the Responsible Person level if your CPSR (Cosmetic Product Safety Report) doesn’t include clinical or instrumental support
TL;DR: Knowing which lane you’re in determines whether you need a 30-subject repeat insult patch test (RIPT) or a full 60-subject clinical with TEWL and Corneometer endpoints — and that’s a cost and timeline difference worth planning around
Key Technical Parameters #
Sensitive skin claims feel simple on a product brief. In practice, getting a single SKU compliant across the EU, US, and China simultaneously is one of the more document-intensive projects we run. The regulatory challenge isn’t just ingredient restrictions — it’s that “barrier repair” and “sensitive skin” are interpreted differently by each market’s authority, which changes what your dossier needs to contain, which claims you can actually print on pack, and which tests the retailer will ask for before they list you. Brands that launch in one market first and retrofit for others almost always hit delays. This guide maps the compliance requirements as we work through them internally, from initial formulation sign-off through to shelf-ready documentation, so you can plan the burden before you brief us.
What Actually Determines Your Compliance Burden — Before the Formulator Opens a Database #
The first question we ask when a brand brings us a barrier repair brief isn’t “what actives do you want?” It’s “which markets, in which order, and what claims are you committing to on pack?”
That sequence matters more than most teams anticipate. “Barrier repair” is not a regulated claim in the US — the FDA Cosmetics Guidelines treat it as a cosmetic claim that requires no pre-market approval, just substantiation if challenged. In the EU, the same phrase sits in a grey zone under EU Cosmetics Regulation 1223/2009, where it’s permissible as a cosmetic claim but will trigger tighter scrutiny at the Responsible Person level if your CPSR (Cosmetic Product Safety Report) doesn’t include clinical or instrumental support. In China under the NMPA Cosmetic Regulation, “barrier repair” has been categorized under special cosmetics provisions since the 2021 regulatory overhaul for some functional formats, which means you may need registration rather than filing — a difference of 6 to 12 months in lead time.
Get the market sequence wrong and you’re reformulating mid-project. We’ve seen that happen with otherwise well-run brand launches.
The claim also shapes the testing stack. A product positioned as “for sensitive skin” with no therapeutic implication needs a different dossier than one claiming to “repair the skin barrier” with before-and-after TEWL data. Knowing which lane you’re in determines whether you need a 30-subject repeat insult patch test (RIPT) or a full 60-subject clinical with TEWL and Corneometer endpoints — and that’s a cost and timeline difference worth planning around.
Compliance Requirements by Market — Head-to-Head #
Below is how we structure the compliance matrix internally when a brand is targeting multi-market launch. The criteria reflect what we’re actually asked to produce, not just what the regulations say in theory.
| Requirement | EU (Reg 1223/2009) | US (FDA / Voluntary) | China (NMPA 2021) |
|---|---|---|---|
| Pre-market approval | No (notification via CPNP) | No | Yes — registration for special cosmetics; filing for general |
| Safety assessment format | Full CPSR (Part A + B), toxicologist sign-off | No mandated format; FTC substantiation standard applies | Safety assessment per NMPA template; local testing lab required |
| “Barrier repair” claim | Permissible as cosmetic claim; CPSR must include efficacy substantiation | Permissible; no mandated test, but retailer SOP may require dermatest data | May trigger special cosmetic classification depending on format and wording |
| Fragrance allergen disclosure | 26 EU allergens must be listed if above threshold (0.001% leave-on) | Not mandated federally; California AB-2762 partial requirement | Not currently mandated, but evolving |
| Preservative restrictions | Annex V positive list; concentration limits per substance | Broad permissibility under GRAS / CIR; no positive list | Positive list system, broadly aligned with EU but with differences in permitted levels |
| Dermatologist-tested claim | Requires substantiation; format not prescribed | Requires substantiation; FTC governs against misleading claims | Requires clinical support from Chinese-accredited institution |
| Key testing minimum | RIPT or HET-CAM / EpiDerm for irritation; SPF if claimed | RIPT standard for “hypoallergenic”; clinical as needed | RIPT equivalent conducted in China; microbial challenge test required |
| Typical dossier preparation time | 6–10 weeks (first-time submission) | 3–5 weeks (no government filing) | 4–9 months depending on classification |
The gap that catches brands off-guard is the China timeline. Six months minimum for registration-track products isn’t uncommon even with a clean formulation. If you’re planning a simultaneous global launch, the NMPA clock needs to start before you finalize the EU CPNP notification — not after.
For EU, the one we flag at every kickoff is the fragrance allergen threshold on leave-on products. The 0.001% (10 ppm) disclosure limit for the 26 identified allergens under Regulation 1223/2009 Annex III applies to rinse-off products at 0.01%. That’s a 10-fold difference. We’ve had brands brief us on a “clean and minimal” barrier cream that still had six listable allergens from the botanical complex they wanted to include. Reformulation was needed. The fix wasn’t difficult, but it added three weeks.
For US, the practical compliance driver in sensitive skin is increasingly retailer-led rather than regulatory. Retailers like Sephora, Target, and specialty pharmacy chains have their own “clean” or “sensitive skin” standards that go beyond FDA requirements. These vary and change without formal notice. Our recommendation is to confirm the specific retailer standard at brief stage — not after the formula is locked.
The Overlooked Variable: Testing Lab Jurisdiction #
Standard comparisons focus on ingredient lists and claim language. The one factor that derails multi-market projects more consistently, in our experience, is testing lab jurisdiction.
For China registration, testing must be conducted at an NMPA-accredited Chinese testing institution. Results from a European or US RIPT study are not accepted as substitutes. This is non-negotiable, and no amount of documentation from a globally recognized CRO changes it. We work with two qualified labs in Shanghai and Guangzhou. First-time brands often don’t budget for duplicate testing, which creates a cost and timeline shock mid-project.
For EU, the CPSR safety assessment must be signed by a qualified toxicologist meeting the criteria in Annex I of Regulation 1223/2009 — a recognized qualification in pharmacy, medicine, toxicology, or similar, with the relevant degree. Not every safety assessor your brand has used historically will meet this qualification in every EU jurisdiction. We’ve had two projects where a Responsible Person’s toxicologist didn’t qualify under the stricter interpretation applied in Germany and the Netherlands, requiring reassessment.
There’s also the question of clinical data provenance. A 2022 split-face RCT (n=48, 12 weeks) evaluating a ceramide-dominant barrier repair emulsion found a 34% reduction in TEWL versus vehicle control, measured via Tewameter TM 300. Strong data. But the study was conducted in South Korea, and when one client tried to use it for NMPA registration, the accreditation wasn’t accepted. The data was solid. The jurisdiction was wrong. We use this case in our internal QA-11 compliance intake form as a reminder to verify study origin before it becomes part of the registration package.
For clinical data to support EU CPSR and US retailer requirements simultaneously, a single well-designed study run at a European CTFA-aligned CRO with an n=40 or above will generally satisfy both. That’s the efficient path for brands that aren’t pursuing China in the same cycle.
Different groups have different practices here. Some contract manufacturers push for centralized testing using one European lab for everything, then handle China separately. Others use Chinese CROs for all testing on the logic that the data will be accepted everywhere except the EU, requiring only a top-up study for CPSR. Our practice is to recommend the European-first approach for brands whose primary market is EU or US, and Chinese-first for brands leading with China, then planning westward expansion. Neither is universally better — it depends on which market’s timeline is the harder constraint.
Documentation Checklist and Incoming Inspection Priorities #
Once the compliance path is defined, the documentation build runs in parallel with formulation development. Below is what we track through our QA-11 compliance intake form for a typical barrier repair leave-on product targeting EU + US:
Minimum documentation stack, EU + US launch:
- Full ingredient INCI list with CAS numbers and supplier-specific grade
- Safety data sheets for each raw material (current revision, dated within 3 years)
- Allergen declaration from fragrance supplier confirming levels per EU Annex III thresholds
- Stability study data: 12 weeks at 40°C/75% RH minimum (ICH-aligned, see ICH Stability Guidelines), 25°C/60% RH as long-term, with freeze-thaw cycling
- Preservative efficacy test (PET) per ISO 11930, repeated on final commercial formulation, not prototype
- Compatibility data for final packaging material
- RIPT study (minimum 50 subjects, 3-week repeat insult protocol) or HET-CAM/EpiDerm for irritation potential
- Dermatologist-tested statement substantiation if claimed (signed assessment, methodology documented)
- CPSR Part A (product information) and Part B (safety assessment by qualified toxicologist)
- GMP certificate of the manufacturing site (ISO 22716 compliance)
The incoming inspection priorities when we receive actives from our supplier network for a sensitive skin formulation follow our Category B classification: full CoA verification, microbial limits testing on each incoming lot, and pH verification of water-phase ingredients before they enter production. On barrier repair projects specifically, we also verify ceramide ratio consistency lot-to-lot, because ceramide purity variation across suppliers is wider than the datasheets suggest. We’ve measured up to a 12% variance in ceramide NP content between lots from the same supplier over an 18-month period.
For timeline planning: stable, validated formulation to CPNP notification in the EU typically runs 14 to 18 weeks for a first-time submission with a new Responsible Person. US launch, absent retailer-specific requirements, can move in 8 to 10 weeks from formula lock. NMPA registration, special cosmetics track, is 6 to 12 months from submission. These assume the testing is proceeding in parallel, not sequentially.
One ongoing issue we haven’t fully resolved: there’s no clean alignment between what the SCCS Scientific Opinion considers sufficient for sensitive skin claim substantiation and what major EU retailers are currently asking for. The regulatory floor and the commercial floor are diverging. Our current position is to use the retailer’s requirements as the ceiling and design upward from there — but that may mean over-investing in clinical data for some markets.
Formulation Notes for Brand Partners #
When you brief us on a barrier repair product, the first thing we need is market and claim clarity — not an ingredient list. Which markets are confirmed for launch, and in which order? What’s the on-pack claim language you’re committed to? Is “barrier repair” a cosmetic claim or are you positioning closer to a therapeutic benefit? Those answers shape the entire compliance workstream before we touch the formula.
The mistake we see most often is brands locking a formula before the claim strategy is finalized, then discovering the chosen actives create a documentation burden they hadn’t budgeted for. One recent example: a client requested a formula with a botanical complex and a “repair” claim for EU launch. Three of the botanical extracts carried reportable allergens above the 0.001% threshold. Reformulation added three weeks and a partial re-stability run.
On timeline: once brief is confirmed, lab samples take 2 to 3 weeks. Accelerated stability (40°C/75% RH, ICH-aligned) runs 4 to 8 weeks. Real-time 24-month stability starts concurrently from the first production batch. RIPT and clinical testing run in parallel with stability — typically 8 to 10 weeks for a standard RIPT panel. For China registration, NMPA testing needs to start as early as possible; it will not compress to fit a launch window.
Frequently Asked Questions #
We want to say “repairs the skin barrier” on our EU packaging — is that a problem?
A: As a cosmetic claim, it’s permissible under Regulation 1223/2009, but your CPSR needs substantiation behind it — instrumental data (TEWL, Corneometer) or a clinical study. Without that, a Responsible Person will be reluctant to sign off, and some EU markets’ notified bodies will query the dossier.
Our formula was already tested in a US lab — can we use that data for China registration?
A: Not directly. NMPA requires testing to be conducted at an accredited Chinese institution. US or European RIPT results don’t substitute. Budget for duplicate testing if China is in scope — it’s not optional and it adds time regardless of how clean the formula is.
What’s the most common reason stability fails on barrier repair products during the compliance process?
A: Preservative efficacy failure on the final production formula, after it passed on the prototype. When packaging changes between prototype and commercial — cap liner, pump material, jar coating — it can introduce microbial or chemical incompatibility. We always require a PET repeat on the final commercial formulation, not just the lab batch. Across our projects, roughly one in four products that passed PET at prototype stage needed adjustment after packaging was finalized.
What’s a realistic minimum order quantity and timeline for a new barrier repair SKU?
A: MOQ on a custom formula is typically 500 kg for cream or lotion formats, or 1,000 units for smaller fill weights in a dedicated run. Timeline from brief confirmation to first production sample is 4 to 6 weeks; from sample approval to first commercial production is an additional 6 to 8 weeks, assuming compliance documentation is running in parallel. If NMPA registration is in scope, the overall timeline extends considerably — plan for 9 to 14 months from brief to China shelf.
Should we pursue ISO 22716 GMP certification from our manufacturing partner, or is it enough that they self-declare?
A: Self-declaration isn’t sufficient for EU compliance — your Responsible Person needs a GMP certificate or audit evidence. ISO 22716 is the recognized standard; a third-party audit certificate carries more weight with EU notified bodies than a self-assessment document. For US retail specifically, some pharmacy and specialty retail buyers now require it as a condition of listing, not just a nice-to-have. Verify this requirement with your target retailers before you assume self-declaration will clear the procurement process.
Have a product concept in mind? Contact our formulation team to request a complimentary brief review.
The TEWL endpoint piece is where we’ve burned the most time — our first 60-subject clinical for a ceramide-dominant barrier cream (launched 2022, Seongnam R&D site) came back with Corneometer data that satisfied EU CPSR reviewers but got flagged by a major US e-tailer whose internal SOP wanted a separate 28-day RIPT from a US-accredited lab, which wasn’t in the original study design at all.
We launched a ceramide-niacinamide barrier cream across EU and China simultaneously and didn’t realize our “barrier repair” wording would trigger special cosmetic classification with NMPA until we were already three months into CPNP notification prep. Our China filing had to be completely restructured, local lab retesting added another 14 weeks, and we missed the Q4 retail window entirely. Briefed the OEM with one claim set assuming they’d flag the divergence — they didn’t.
The NMPA “special cosmetic” trigger is the one that catches brands off guard most often — we’ve seen “barrier repair” wording get flagged in China even when the exact same claim sailed through CPNP notification in the EU with just instrumental TEWL data attached to the CPSR. Japan adds another wrinkle: quasi-drug classification under MHLW can kick in depending on how you phrase the mechanism, and that’s a full additional approval track most timelines don’t budget for.
We learned this the hard way launching our barrier cream into Sephora EU and China in the same quarter — the CPSR sign-off came back clean but our NMPA filing stalled for four months because our instrumental data (TEWL + Corneometer) was collected at a non-recognized local lab. Run your efficacy testing at a China-accepted facility from the start, even if you’re EU-first.
The retailer SOP piece doesn’t get enough attention — we had Dermstore request a full dermatest report on a ceramide serum we’d already sold into Whole Foods for 18 months without any documentation issues.
The FTC substantiation standard for US claims sounds permissive until you try to run it backwards from a specific claim — “restores skin barrier in 7 days” on a ceramide product will get you a competent and reliable scientific evidence demand fast, and that standard in practice often lands you at a 30-subject RIPT minimum anyway, just without anyone telling you upfront that’s where you’d end up.
Oat beta-glucan has been our go-to for sensitive skin positioning since 2021, and the molecular weight variation between suppliers is something we still don’t have a clean answer on — high-MW batches consistently outperform on our TEWL readings but the supplier COAs rarely specify anything beyond purity percentage. We’ve had two reformulations stall at CPSR stage because our toxicologist flagged inconsistent characterization data across lots.
When you’re doing simultaneous EU/US/China launches and the OEM is running clinical substantiation for the CPSR, do you require that the 60-subject TEWL study be conducted at a lab with both ISO 17025 accreditation and NMPA-recognized status, or does the NMPA safety assessment get supported by a separate China-side study entirely?
Retrofitting for China after EU launch is where we’ve consistently lost the most calendar time — our 2023 ceramide-cholesterol-fatty acid barrier serum (3:1:1 ratio, formulated at our Seoul contract site) had a clean CPSR and was live in CPNP before we realized the cholesterol concentration we’d used to hit the clinical efficacy numbers was going to complicate our NMPA general cosmetic filing in a way the EU dossier gave us zero signal about.
Japan’s quasi-drug (医薬部外品) classification is the one that blindsides most brands after they’ve cleared EU and China, because ingredients like allantoin and dipotassium glycyrrhizate that sail through CPNP as standard cosmetic actives can push a formula into quasi-drug territory with PMDA, which adds a 12-18 month approval track that nobody budgeted for. We had a ceramide-allantoin barrier serum ready for Cosme Kitchen in Q3 2023 and basically had to choose between reformulating or delaying Japan launch by over a year.