Writing a Cosmetic Product Safety Report (CPSR): Part A and Part B Explained

Under Article 10 and Annex I of EU Regulation (EC) No 1223/2009, every cosmetic product placed on the EU market must have a Cosmetic Product Safety Report (CPSR). The CPSR is the safety dossier that sits inside the Product Information File (PIF), demonstrating that the product is safe for human health when used under reasonable and foreseeable conditions.

The CPSR is split into two parts: Part A — Safety Information and Part B — Safety Assessment. Each has prescribed content and prescribed authorship. This guide walks through what each part requires and how to structure a CPSR that holds up under competent-authority inspection.

Who can write a CPSR

Part A can be assembled by any technically competent person — typically a regulatory affairs specialist, formulator, or contract toxicologist. The data in Part A is largely documentary: ingredient compositions, stability test results, microbiological challenge test results, exposure estimates.

Part B must be written by a qualified Safety Assessor. Article 10(2) defines this: a person holding a diploma from a university course in pharmacy, toxicology, medicine, or a similar discipline, or a course recognized as equivalent by a Member State. The Safety Assessor must sign Part B and include their qualifications.

Many small and mid-size brands contract Safety Assessors externally. Hourly rates of €150-450 are typical for EU Safety Assessors; a full CPSR for a single product runs 8-20 hours of Safety Assessor time depending on complexity.

Part A — Safety Information

Annex I, Part A enumerates ten subsections:

1. Quantitative and qualitative composition

The complete formulation: every ingredient by INCI name, CAS and EC numbers where applicable, function, and exact percentage. Note: the CPSR uses the exact composition, not the frame formulation submitted to CPNP. Trade-name flavors and fragrances must be broken down to the regulated constituent level (especially for the 26 EU fragrance allergens above 0.001% leave-on / 0.01% rinse-off).

2. Physical and chemical characteristics and stability

Stability data: typically 3-month accelerated stability at 40°C plus real-time stability at 25°C, with measurements of pH, viscosity, color, odor, and microbiological status. Cosmetics with predictable stability windows (e.g. well-characterized creams) may use shorter studies; novel formats need longer.

3. Microbiological quality

Challenge test results (e.g. ISO 11930) demonstrating the preservation system suppresses microbial growth under contamination scenarios. Specifications for finished product microbial limits.

4. Impurities, traces, information about the packaging material

Identified impurities from raw materials (e.g. nitrosamines from amino-functional ingredients, 1,4-dioxane from ethoxylates, residual solvents). Packaging material composition and migration testing if relevant.

5. Normal and reasonably foreseeable use

Where on the body, frequency of application, duration of contact, leave-on or rinse-off, target population (e.g. children under 3 trigger different rules). Also includes "foreseeable misuse" — eye contact for a face cream, accidental ingestion for a lip product.

6. Exposure to the cosmetic product

Quantitative exposure estimate in mg/kg body weight per day. Standard SCCS Notes of Guidance provide default exposure values per category (e.g. body lotion = 7.82 g/day, face cream = 0.6 g/day applied to defined body areas). Calculate aggregate exposure if multiple products in the same line use the same ingredient.

7. Exposure to the substances

For each ingredient, the systemic exposure dose (SED) in mg/kg/day, derived from the product exposure and the ingredient concentration. Compare against the No Observed Adverse Effect Level (NOAEL) to calculate Margin of Safety (MOS).

8. Toxicological profile of the substances

For each ingredient, the toxicological data: acute toxicity, repeated-dose toxicity, sensitization, irritation, genotoxicity, carcinogenicity, reproductive toxicity, photoinduced toxicity. Cite SCCS opinions, CIR safety reviews, ECHA REACH dossiers, or published primary literature.

9. Undesirable effects and serious undesirable effects

Data on undesirable effects from clinical safety testing, post-market surveillance, and adverse events. Compare against expected effects for the product category.

10. Information on the cosmetic product

Any additional data — clinical efficacy studies, consumer perception studies, comparable-product data — relevant to the safety assessment.

Part B — Safety Assessment

Part B is the qualified Safety Assessor's evaluation of Part A. Annex I, Part B enumerates four subsections:

1. Assessment conclusion

The Safety Assessor's overall conclusion: is the product safe for human health under normal and reasonably foreseeable use? This is a yes/no statement with supporting rationale.

2. Labelled warnings and instructions of use

The Safety Assessor identifies whether the product requires any specific label warnings (e.g. "avoid contact with eyes," "do not use on children under 3," allergen disclosures). Compare against legally required warnings in Annex III and assessor judgment for residual risks.

3. Reasoning

The detailed reasoning that supports the conclusion. This is the heart of Part B — it ties together exposure data, toxicological data, and the assessor's professional judgment. It must reference Part A data and cite the SCCS Notes of Guidance and any authoritative opinions used in the assessment.

4. Assessor's credentials and approval

Name, qualifications (diploma, university, date), professional title, signature, and date. This is the assessor's accountability statement.

Common CPSR mistakes

• Using a template without a real Safety Assessor. Part B requires a qualified human's judgment. Competent authorities will check the assessor's credentials during inspection. Templates filled in by non-qualified staff don't satisfy Article 10.
• Frame formulation in Part A. Part A requires the exact composition, not the CPNP frame. Brands sometimes submit the frame in both, leading to a CPSR that doesn't match what's in the bottle.
• Missing aggregate exposure. If multiple products in a regimen contain the same ingredient, aggregate exposure must be considered. A vitamin C serum + vitamin C moisturizer + vitamin C eye cream regimen contributes more ascorbic acid exposure than any single product.
• Stale toxicological data. SCCS opinions are updated regularly. Citing a superseded opinion looks bad and risks inspection findings. Re-check at each CPSR review cycle (typically annually).
• No documented review on reformulation. If you change an ingredient or its concentration, Part A must be updated AND Part B must be re-assessed by the Safety Assessor. A new sticker won't do.

How Cosmetica supports CPSR work

Cosmetica's PIF Builder:

• Assembles Part A from your ingredient list, stability test imports, and packaging data.
• Pre-fills toxicological profiles from a curated library of SCCS opinions, CIR safety reviews, and published literature — every cited ingredient links back to its primary source.
• Calculates exposure (SED) and Margin of Safety (MOS) automatically using SCCS Notes of Guidance defaults, which the Safety Assessor can override.
• Outputs a structured Part A draft for a qualified Safety Assessor to review and sign Part B against — saving hours of formatting and citation lookup.

Cosmetica doesn't replace the Safety Assessor — by EU law, that's a human qualified individual. What it does is hand the Safety Assessor a well-structured Part A so they can spend their time on judgment rather than typesetting.

See Cosmetica's EU CPNP coverage → | What is a CPSR? →