Hypoallergenic & Fragrance-Free Formulation: Allergen-Free Ingredient Selection & Patch Test Protocol

Overview #

Fragrance is the single leading cause of cosmetic contact allergy in the EU, accounting for roughly 1–4% of patch-tested dermatology patients reacting to the fragrance mix alone. That number sounds small until you’re a brand trying to sell into sensitive-skin retail channels where one complaint thread can kill a launch. Our position at the bench is straightforward: “fragrance-free” and “hypoallergenic” are not marketing decisions — they are formulation architecture decisions that have to be made at the ingredient selection stage, not bolted on at the end. This article walks through the clinical evidence behind the actives we actually use in these formulations, how we run patch test protocols at scale, and what claim substantiation looks like across EU, US, and NMPA markets.

The Allergen Problem Is Bigger Than Fragrance #

Most brand briefs we receive say “fragrance-free” and stop there. That’s the easy part. The harder conversation is about the 26 EU-listed fragrance allergens that can appear in ingredient categories brands don’t think of as fragrance — essential oils, botanical extracts, certain preservatives, even some emulsifiers. Under EU Cosmetics Regulation 1223/2009, any of the 26 listed allergens present above 0.001% in leave-on products must be declared on-pack. The 2023 amendment extended this list significantly, and we’ve had to reformulate several SKUs mid-development because a botanical extract the brand loved was carrying linalool or limonene above threshold.

On our production line, the failure mode we see most often is a brand approving a “natural” formula that contains lavender oil or rose extract, then being surprised when the INCI declaration triggers a retailer’s allergen screening. We now require suppliers to provide full GC-MS data on any botanical extract before it enters our approved raw material list. Not a certificate of analysis. Actual chromatography.

The preservative question is equally thorny. Methylisothiazolinone (MI) was restricted to rinse-off only in the EU at 0.0015% after a wave of sensitization cases. We stopped using it in leave-on formats entirely in 2019 — not because regulation forced us to at the time, but because the SCCS Scientific Opinion data was already pointing that direction. Our current leave-on preservative systems for sensitive-skin formulas run on phenoxyethanol at ≤0.8% combined with ethylhexylglycerin, or sodium benzoate/potassium sorbate dual systems buffered to pH 5.0–5.5. Both pass our 28-day challenge test (ISO 11930) consistently.

Clinical Evidence for Key Barrier-Repair Actives #

This is where we spend most of our formulation time on sensitive-skin briefs. Three actives dominate our evidence-based toolkit: ceramides, colloidal oatmeal, and niacinamide. Each has a meaningfully different evidence profile, and brands often conflate them.

Ceramides

The strongest mechanistic data we work from is a double-blind, randomized controlled trial (n=49, 8 weeks) that measured transepidermal water loss (TEWL) reduction in subjects with mild-to-moderate atopic dermatitis using a ceramide-dominant emollient twice daily. TEWL decreased by 34% versus 11% in the vehicle control group. Stratum corneum hydration improved by 28% on the Corneometer scale. What that study doesn’t capture — and what we’ve learned from our own batches — is that ceramide ratio matters as much as ceramide concentration. The skin’s natural ratio runs approximately 1:2:1 for ceramide:cholesterol:free fatty acid. Formulas that load ceramide NP at 2% without the cholesterol and fatty acid co-factors don’t replicate that ratio, and in our internal stability and skin feel assessments, they perform noticeably worse on barrier recovery endpoints.

We use a pseudo-ceramide or phytoceramide blend at 0.5–1.5% combined with cholesterol (0.3%) and linoleic acid (0.2%) in most of our barrier-repair cream formats. That’s the architecture. The exact percentages shift depending on the base emulsion system.

Colloidal Oatmeal

Colloidal oatmeal is one of the few cosmetic actives with an FDA OTC monograph behind it — specifically for skin protectant claims at 0.003–3.0% concentration. A randomized, double-blind, vehicle-controlled study (n=58, 12 weeks) in subjects with mild atopic dermatitis showed a 32% reduction in IGA (Investigator’s Global Assessment) score versus 18% for vehicle. Itch severity scores dropped by 41% in the active arm. The active compound driving most of this is avenanthramides, a group of phenolic alkaloids with documented anti-inflammatory and antipruritic activity.

Practically speaking, colloidal oatmeal is also one of the more forgiving actives to work with from a stability standpoint. It doesn’t require low pH, doesn’t interact badly with most preservative systems, and doesn’t destabilize emulsions at typical use levels. For brands targeting eczema-prone or reactive skin, it’s often our first recommendation. See our broader approach to barrier repair and sensitive skin formulation for how we layer this into full product architectures.

Niacinamide

The niacinamide evidence base is genuinely strong, and we use it at 4–5% in most sensitive-skin serums and moisturizers. A split-face, double-blind RCT (n=50, 8 weeks) showed 5% niacinamide reduced sebum excretion rate by 19% and improved skin barrier function (measured by TEWL reduction) by 24% versus vehicle. A separate 12-week study (n=44) demonstrated a 20% reduction in erythema index scores in rosacea-prone subjects.

One thing we’re honest with brands about: the niacinamide-vitamin C interaction story is more complicated than the internet makes it sound. At typical use concentrations and modern formulation pH ranges, the niacin flush concern is largely theoretical. We’ve run compatibility testing on combined niacinamide + ascorbic acid formulas at 5% + 10% respectively, and we don’t see the yellowing or niacin formation that older literature predicted — provided pH stays between 5.5 and 6.5 and the formula doesn’t sit at elevated temperature for extended periods. For more on vitamin C system design, see our vitamin C and antioxidant systems technical documentation.

Evidence Strength Comparison: Barrier-Repair Actives #

Honestly, allantoin is on almost every sensitive-skin brief we receive, and the evidence base is the weakest of the group. It’s not that it doesn’t work — it’s that the clinical data is thin and old. We use it, but we don’t build claims around it.

Patch Test Protocol: What We Actually Run #

A lot of brands ask us about patch testing as if it’s a single test. It’s not. There are at least three distinct protocols relevant to a hypoallergenic claim, and conflating them is where projects go sideways.

Repeat Insult Patch Test (RIPT) is the standard for sensitization potential. We run it with a minimum of 50 subjects (panelists with self-reported sensitive skin), 9 induction applications over 3 weeks, a 2-week rest period, and a challenge application. A product claiming “hypoallergenic” in the EU context should have zero sensitization reactions at challenge. We’ve had one formula fail RIPT in the last three years — a botanical-heavy serum where the supplier changed the extraction solvent on a chamomile extract mid-project without notifying us. The new extract carried a higher bisabolol oxide A load. We caught it at RIPT, not at stability. That’s why we re-test whenever a raw material source changes.

Human Repeat Insult Patch Test (HRIPT) is essentially the same protocol but with a larger panel — typically 100–200 subjects for stronger claim support. For brands targeting clinical retail or pharmacy channels, we recommend HRIPT over RIPT. The cost difference is real but manageable.

Dermatologist-tested claims require a supervised use study, typically 4 weeks, with dermatologist assessment at baseline and endpoint. This is the most common claim we support for EU and US markets. It doesn’t require zero reactions — it requires dermatologist sign-off that the product is suitable for sensitive skin.

One thing we push back on regularly: brands wanting to run patch testing on a formula that hasn’t been stability-tested first. We’ve seen formulas pass RIPT at T=0 and then generate irritation complaints in market because the preservative system degraded by month 3. Stability first. Always.

Where Most Brands Get the Claim Substantiation Wrong #

EU, US, and NMPA have meaningfully different frameworks here, and the gap between them is wider than most brand owners realize.

Under EU Cosmetics Regulation 1223/2009 and the associated Claims Regulation (EC) No 655/2013, “hypoallergenic” is a claim that must be substantiated with documented evidence. There is no legal definition of hypoallergenic in EU law, which sounds permissive but actually creates liability — if a consumer challenges the claim, you need a dossier. We build that dossier to include: full allergen screening against the 26 listed substances, RIPT or HRIPT data, and a Product Information File (PIF) with the safety assessment signed by a qualified safety assessor.

In the US, FDA Cosmetics Guidelines take a different position: “hypoallergenic” has no regulatory definition and FDA does not require pre-market approval for cosmetic claims. That sounds easier. In practice, the FTC’s substantiation standard still applies — claims must be truthful and backed by competent and reliable scientific evidence. For a US brand, RIPT data plus a dermatologist-tested study is the practical minimum for defensible claims.

NMPA is the most prescriptive of the three. Under NMPA Cosmetic Regulation, “sensitive skin suitability” claims require a human safety test conducted by a NMPA-recognized testing institution, following the Technical Guidelines for Human Safety Testing of Cosmetics. The test panel must include ≥30 subjects with confirmed sensitive skin. Results must be filed as part of the registration or notification dossier. We’ve run these tests for multiple brand partners entering the China market, and the timeline adds roughly 8–12 weeks to the development schedule. Plan for it.

One area where we’re still not fully convinced the regulatory frameworks have caught up: microbiome-related claims on sensitive skin products. “Microbiome-friendly” is appearing on packaging everywhere, but the substantiation standards are inconsistent across markets. This is still evolving.

Formulation Notes for Brand Partners #

What market? What are you expecting on-pack? Those are the first two questions we ask when a sensitive-skin brief lands on our desk.

If you’re targeting EU pharmacy or clinical retail with a “hypoallergenic, dermatologist-tested” positioning, the formulation architecture needs to be locked before testing begins — because any ingredient change after RIPT invalidates the data. That means your fragrance-free, allergen-screened ingredient list needs sign-off before we start stability. We typically run a 4-week accelerated stability study (40°C/75% RH) in parallel with RIPT to compress the timeline.

If you’re targeting the US mass or prestige channel, the regulatory bar is lower but the retailer bar is often higher. Sephora’s clean standard, Target’s clean list, and Ulta’s guidelines each have their own restricted ingredient lists that go beyond EU regulation in some areas. We maintain an updated restricted ingredient database cross-referenced against major US retailer standards — it saves significant reformulation time.

For NMPA registration, the sensitive skin claim pathway requires us to use only NMPA-registered raw materials. That limits some of the newer ceramide and peptide ingredients that haven’t yet completed China registration. We flag this early in the brief process.

Budget reality: a full hypoallergenic claim package — allergen screening, RIPT (n=50), dermatologist-tested study, stability — runs approximately $8,000–$15,000 USD in third-party testing costs depending on the testing institution and market. That’s before formulation and production. Most indie brands underestimate this. We’d rather tell you upfront than have the conversation at launch.

Frequently Asked Questions #

Q: Can we just say “fragrance-free” without running patch tests?
Yes, technically — “fragrance-free” is a compositional claim, not a performance claim, so it doesn’t legally require patch test data in most markets. But if you’re also claiming “suitable for sensitive skin” or “hypoallergenic,” you need the testing. In our experience, brands that launch fragrance-free without any patch data are one bad review away from a claim challenge.

Q: We want to use a natural rose extract for marketing — can we keep the fragrance-free claim?
Almost certainly not. Rose extract typically carries geraniol, citronellol, and eugenol — all on the EU 26 allergen list. At the concentrations needed for a meaningful marketing story, you’ll breach the 0.001% declaration threshold for leave-on products. We’ve had this conversation with at least a dozen brands. The answer is usually: pick the claim or pick the extract.

Q: What’s the minimum panel size for a credible hypoallergenic claim?
For RIPT, 50 subjects is the practical minimum for EU and US markets. For NMPA sensitive skin claims, the guideline specifies ≥30 subjects with confirmed sensitive skin. We recommend 50 regardless of market — it gives you a single dataset that works across all three regulatory environments.

Q: How long does the full patch test and claim substantiation process take?
RIPT alone runs 5–6 weeks from panel recruitment to final report. Add a 4-week dermatologist-tested study running in parallel, and you’re looking at 6–8 weeks total for the testing phase. NMPA-specific human safety testing adds another 8–12 weeks if you need China registration. Build this into your launch timeline from day one — it cannot be compressed.

Q: Our formula passed RIPT — can we now change the fragrance to a “natural” alternative?
No. Any ingredient change after RIPT invalidates the test data for that specific formula. You’d need to re-run. This is the most common and most expensive mistake we see in sensitive-skin development. Lock the formula. Test it. Then don’t touch it.

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