Stability Testing Your CPSR Needs Before EU Launch

A safety assessor asked a client of mine for stability data on a whipped body butter last year, and the maker's honest response was "it's been in my kitchen for four months and it still smells fine." That is not stability data. That is an anecdote. The EU's Cosmetic Product Safety Report needs something a lot more structured before it can be signed off, and stability testing is where a surprising number of otherwise-ready launches stall.

Why stability sits inside the CPSR conversation

Under Regulation (EC) No 1223/2009, placing a cosmetic on the EU market means a Responsible Person established in the EU keeps a Product Information File, and inside that PIF sits the Cosmetic Product Safety Report, signed by a qualified safety assessor. The assessor is not just checking your ingredient concentrations against restricted lists. They are forming a judgment about whether the product stays safe and performs as claimed for its intended shelf life, under realistic storage and use conditions. Stability testing is the evidence that judgment rests on.

Skip stability data and you are asking an assessor to sign off on durability claims with nothing to back them. Most will not, and they shouldn't.

The two testing tracks

Accelerated stability testing puts the product under exaggerated stress, elevated temperature, temperature cycling, light exposure, sometimes humidity, over a short window, typically weeks rather than months. The idea is to force degradation processes to show up faster than they would at normal storage conditions, giving you an early read before you have real-time data to lean on.

Real-time stability testing stores the product under normal, expected conditions and observes it over the actual intended shelf life, or long enough to support the claim you want to make. This is slower but it is the data that ultimately holds the most weight, since it reflects what actually happens rather than an accelerated proxy.

In practice, most brands run both. Accelerated data lets you move toward launch without waiting a full year, while real-time testing runs in parallel and eventually confirms or corrects what the accelerated data suggested.

What assessors are actually watching for

• Physical stability: separation, crystallization, color drift, texture change, viscosity shift.
• Chemical stability: whether actives or preservatives degrade below effective levels over time.
• Microbiological stability: whether the preservative system continues to control microbial growth through the claimed shelf life, often assessed alongside a separate challenge test.
• Packaging compatibility: whether the formula interacts with its container, for instance a fragrance oil that degrades a certain plastic, or a low-pH product that corrodes a metal component.

How this feeds the durability claim

Your shelf life claim, whether that is an expiry date or a Period After Opening symbol, needs to be something the stability data actually supports. If your real-time data only covers eight months and you want to claim a two-year shelf life, that gap needs to be bridged with accelerated data and a documented rationale the assessor is comfortable defending, not just an optimistic guess based on how the product looked last time you checked the jar.

A practical protocol outline

Stage	Typical approach	What it tells you
Initial baseline	Full characterization at time zero	Reference point for all later comparisons
Accelerated stress	Elevated temperature cycling over several weeks	Early signal of instability before real-time data matures
Real-time storage	Ambient and recommended storage conditions over the intended shelf life	The data an assessor weighs most heavily
Freeze-thaw cycling	Repeated temperature swings, relevant for emulsions	Emulsion breakdown risk during shipping or storage
Packaging compatibility	Product stored in final packaging, not just lab containers	Interaction between formula and container

Where this leaves a small brand

If you are a solo formulator without an in-house lab, you generally have two workable paths: contract the testing out to a lab that specializes in cosmetic stability work, or work closely with your safety assessor early enough that they can tell you exactly what evidence they need before you are three weeks from a launch date with no data in hand. The second path is cheaper and it is the one most small brands wish they had taken from the start.

Stability data is also not a one-time exercise. Any meaningful reformulation, a new preservative, a changed emulsifier, a different pH target, resets the clock and generally calls for fresh stability work before the PIF can be considered current again.

None of this is something Cosmetic Comply handles directly today, since our live workflow covers the Canadian CNF process, but if EU support comes online down the line, the same principle will apply there as it does here: the paperwork is only as good as the data feeding it, and stability testing is not a box you check quickly at the end.