Cosmetics Stability Testing Under EU Regulation 1223/2009: What Your Product Information File Actually Needs

Every few months, a safety assessor returns a cosmetics safety report with the same annotation: insufficient stability data. The brand assumed their six-month observation at ambient temperature was enough. It wasn’t. Under Regulation (EC) No 1223/2009, stability testing is a documented legal obligation — and the bar for what constitutes acceptable data is higher than most brands realise.

This gap is especially common among smaller manufacturers and indie brands entering the EU market for the first time. They complete the CPNP notification, appoint a Responsible Person, and commission a safety assessment — then hand the safety assessor a one-page lab report covering pH and visual appearance at 25°C. That report won’t anchor a compliant Annex I Part B safety report. And it won’t protect you if the ANSM or DGCCRF launches an inspection.

Here’s what the regulation actually says, and what a defensible stability dossier looks like in practice.

What EC Regulation 1223/2009 Actually Requires

Article 10(1) of Regulation (EC) No 1223/2009 states that a cosmetic product may only be placed on the EU market if it has undergone a safety assessment and a safety report has been drawn up in conformity with Annex I. Annex I, Part B — the Cosmetic Product Safety Report — requires the safety assessor to document “the stability of the cosmetic product under reasonably foreseeable storage conditions.”

That phrase, reasonably foreseeable storage conditions, does a lot of work. It means your testing should reflect how the product will actually be stored and used — not just controlled laboratory conditions. A vitamin C serum sitting in a bathroom cabinet exposed to daily heat and humidity cycles needs different test parameters than a powder stored in a cool, dark warehouse.

The regulation doesn’t prescribe a specific test protocol or number of time points. What it does require is that the safety report include data sufficient for the safety assessor to conclude that the product will remain safe and effective for its stated shelf life. The scientific methodology must be justified and documented. “Ambient observation for 6 months” cited without a defined protocol does not meet that standard.

There’s also a direct labelling consequence. Under Article 19(1)(c), cosmetics with a shelf life of 30 months or less must display a “best before” date. Products with a shelf life exceeding 30 months must instead display the Period After Opening (PAO) symbol — the open-jar icon — indicating how long the product remains safe after first use. Stability testing must underpin whichever claim you make on the label. If you’re labelling a product with a 36-month shelf life and a 12-month PAO, you need data supporting both timelines.

Real-Time vs Accelerated Testing: What Each Can and Cannot Prove

Most brands are familiar with accelerated stability testing. You store the product at elevated temperature and humidity — typically 40°C / 75% relative humidity for 6 months, following Cosmetics Europe guidance — and use the results to model real-time shelf life. The Q10 rule of thumb suggests that a 10°C increase in temperature roughly doubles the reaction rate for many degradation mechanisms. In practice, that means 6 months at 40°C can be used to project approximately 24 months of real-time stability at 20–25°C.

But accelerated testing has well-defined limits. It works reasonably well for hydrolysis and oxidation pathways. It works poorly for physical instability — phase separation in emulsions, for instance, is often driven by specific interfacial dynamics that don’t scale linearly with temperature. And for photodegradation of UV filters, retinol, or fragrance components, accelerated thermal testing tells you nothing at all. Photostability testing requires dedicated UV exposure chambers, and it’s a separate, necessary data set for any product containing photoactive ingredients.

Real-time testing — storing samples at 25°C / 60% relative humidity (or 30°C / 65% RH for products distributed in warmer climates like the Mediterranean region) and monitoring at defined intervals — remains the scientific gold standard. Accelerated data is valuable for early decision-making and for supporting CPNP notifications before your real-time study matures. But for a final safety report that will hold up to scrutiny, most safety assessors and auditors expect to see real-time data, particularly for novel formulations or products with sensitive actives.

The Four Stability Parameters Your PIF Must Document

A stability study that measures only pH and appearance will not satisfy the safety assessor reviewing your Annex I Part B report. These four categories represent the baseline for any EU-compliant stability programme.

1. Physical stability. This covers macroscopic and sensory properties: appearance, colour, odour, texture, viscosity, and — for emulsions — phase separation. Qualitative descriptions (“no visible changes”) are not enough on their own. Viscosity measurements at defined shear rates provide objective, reproducible data. For colour cosmetics and suspensions, particle size distribution over time is also relevant.

2. Chemical stability. This is where most PIF submissions fall short. If your product contains active ingredients — vitamin C (ascorbic acid), retinol, niacinamide, alpha-arbutin, or sunscreen UV filters — their concentration must be measured at each time point against the label claim. HPLC is the standard analytical method. A 10% decline from initial concentration is a common acceptance criterion for functional actives, though the safety assessor may require tighter limits depending on the ingredient’s safety margin and claimed efficacy.

3. Microbiological stability. This is routinely confused with Challenge Testing (Preservative Efficacy Testing, PET), which assesses whether the preservative system can resist deliberate microbial contamination. Stability testing assesses whether microbial counts remain within specification over shelf life under normal storage conditions. Both are required — they answer different questions. The European Pharmacopoeia (Ph. Eur.) 5.1.3 provides the applicable limits: for rinse-off products, the total aerobic microbial count must remain ≤100 CFU/g throughout shelf life; for products intended for the eye area, that limit tightens to ≤10 CFU/g.

4. Packaging compatibility. The container affects the product, and the product can affect the container. Migration of plasticisers or colorants from packaging polymers into the formulation, UV transmission through clear versus opaque containers, and gas permeability affecting oxidation-sensitive ingredients are all stability-relevant factors. For any brand reformulating or switching packaging supplier, updated compatibility data is not optional — it’s part of the Annex I dossier.

Where Stability Dossiers Most Consistently Fail

Based on the Product Information Files we review at Care Europe, three gaps appear with notable regularity.

Insufficient time points. A single measurement at month six is not a stability study — it’s a snapshot. Properly designed studies include a baseline (T=0) and measurements at defined intervals: typically months 1, 3, 6, 9, and 12 for a 12-month accelerated programme. The trend line matters as much as the final result. A product that passes at month 12 but shows a steep degradation curve in months 9–12 signals a potential concern that warrants investigation before commercial launch.

Missing acceptance criteria defined upfront. Every stability protocol must document acceptance criteria before testing begins — not after. If pH must remain between 4.5 and 6.5, that needs to be in the protocol document, not in a retrospective interpretation note. Safety assessors and regulatory inspectors want to see that pass/fail boundaries were established before you saw the data. Post-hoc justification of borderline results is a credibility risk.

No overarching protocol document. Stability data submitted as a collection of individual lab reports, with no framing protocol, is a regulatory red flag. A compliant stability protocol documents: study objectives, sample preparation procedures, storage conditions, test methods (with validated or verified analytical procedures referenced), measurement intervals, acceptance criteria, and the statistical approach for interpreting out-of-trend results. Without this document, even technically sound data is difficult to evaluate and defend.

Building a Programme That Holds Up Over Time

For brands launching a new formulation, the right approach is to initiate both studies simultaneously at formulation sign-off — not after. Start the accelerated study (40°C / 75% RH) at the same time as the real-time study (25°C / 60% RH). The accelerated data supports your initial CPNP notification and interim safety assessment. The real-time study builds over the product’s commercial life and provides the long-term evidence base for any enforcement review or export dossier.

For reformulations — even seemingly minor ones — the scope of repeat testing depends on what changed. A new preservative requires fresh PET data and a new microbiological stability study. A different packaging supplier requires updated packaging compatibility and leachables assessment. A change in fragrance above 0.1% w/w in a leave-on product triggers new sensitisation documentation under Annex I. Small formulation changes rarely have small regulatory consequences.

Document every deviation. If a scheduled time point is missed, note it in the study file and explain why. If a batch used for stability testing isn’t fully representative of commercial production, flag that limitation. The DGCCRF and ANSM don’t just review results during inspections — they review the paper trail that generated those results.

Stability testing done properly takes 12 to 24 months to mature for a new product. That timeline needs to be built into your launch planning from day one — not treated as a compliance checkbox three weeks before your CPNP notification goes in.

---

Written by Nour Abochama, Quality & Regulatory Advisor, Care Europe | VP Operations, Qalitex. Learn more about our team

Talk to our team about EU market entry. Contact us

Related from our network

• ISO 17025-accredited cosmetics and supplement testing — Finished-product stability, microbiological limits, and active ingredient quantification for brands targeting the US market.
• Health Canada NHP compliance and testing support — Canadian regulatory testing and GMP consulting for European brands entering the Canadian natural health products market.