Cosmetic toxicology testing provides the scientific evidence needed to demonstrate that cosmetics and personal care products are safe for their intended use. This technical guide explains the principal toxicology test items—including acute toxicity, skin and eye irritation, sensitization, phototoxicity, genotoxicity, and repeated-dose studies—along with accepted in vitro and in vivo methods, applicable OECD and GB standards, and their role in regulatory registration and product safety assessment.
Technical Overview
Cosmetic toxicology testing is the systematic evaluation of the potential adverse health effects of cosmetic products and their ingredients. The testing battery covers local endpoints such as skin corrosion, skin irritation, and eye irritation, immune-mediated endpoints such as skin sensitization, and systemic endpoints such as acute toxicity, repeated-dose toxicity, reproductive toxicity, and genotoxicity. The resulting data form the scientific backbone of the product safety report and are a prerequisite for market access in most major jurisdictions.
Regulatory frameworks shape the scope of testing. In the European Union, Regulation (EC) No 1223/2009 requires a cosmetic product safety report within the product information file and prohibits animal testing where validated alternatives exist. In China, the Cosmetics Supervision and Administration Regulation (CSAR) and the Safety and Technical Standards for Cosmetics (2015 Edition) define the toxicology studies required for product registration, filing, and new cosmetic ingredient (NCI) dossiers. In the United States, hazard determination under the Federal Hazardous Substances Act framework is supplemented by FDA expectations, while Japan, ASEAN member states, and other markets maintain parallel requirements.
Modern toxicological practice follows a tiered, weight-of-evidence strategy: existing data, quantitative structure–activity relationships (QSAR), and read-across are evaluated first; in chemico and in vitro methods are applied next; and in vivo studies are reserved for cases where no validated alternative is adequate. Integrated approaches to testing and assessment (IATA), including defined approaches for skin sensitization published by the OECD, have substantially reduced reliance on animal models across the cosmetics sector.
Testing Samples
Finished Cosmetic Products
Finished products are typically tested in their marketed form or at their in-use dilution. Leave-on products such as creams, lotions, serums, and sunscreens are tested directly, while rinse-off products such as shampoos and body washes may be diluted to reflect consumer exposure. Special-category products, including oxidative hair dyes and permanent waving agents, are evaluated after mixing with their developer or activator components, since the chemical reaction products can differ markedly from the raw mixture.
Raw Materials and New Cosmetic Ingredients
Ingredient-level testing is mandatory for new cosmetic ingredients and is frequently required for preservatives, UV filters, colorants, botanical extracts, and functional actives. Test substances are characterized for purity, impurity profile, batch consistency, and stability before study initiation, because trace contaminants can materially influence toxicological outcomes.
Sample Preparation and Submission Requirements
- Selection of representative production batches with full batch documentation and certificates of analysis
- Submission of sufficient quantity for all requested endpoints, including reserve for repeat or confirmatory analysis
- Physicochemical characterization covering appearance, pH, water solubility, partition coefficient, and stability
- Vehicle and solvent selection compatible with the biological test system, with concurrent vehicle controls
- Supporting documentation such as the formulation sheet, usage instructions, and safety data sheet
Testing Items
The core toxicology test items for cosmetics fall into well-defined hazard categories, each mapped to validated test guidelines and regulatory expectations.
Acute Toxicity
Acute toxicity studies evaluate adverse effects after a single exposure via the oral, dermal, or inhalation route. Fixed-dose, acute toxic class, and up-and-down procedures (OECD Test Guidelines 420, 423, and 425) have replaced classical LD50 protocols and substantially reduce animal use. Acute oral and acute dermal toxicity tests specified in the Chinese Safety and Technical Standards for Cosmetics remain reference items for certain registration scenarios.
Skin Irritation and Corrosion
Skin corrosion is assessed first using reconstructed human epidermis (RhE) models under OECD Test Guideline 431 or the transcutaneous electrical resistance (TER) assay under OECD Test Guideline 430. Non-corrosive materials are then evaluated for skin irritation using RhE models under OECD Test Guideline 439. The in vivo rabbit patch method (OECD Test Guideline 404) is retained only as a legacy reference where alternatives are unsuitable.
Eye Irritation and Serious Eye Damage
A bottom-up or top-down strategy is applied depending on expected hazard. Bovine corneal opacity and permeability (BCOP, OECD Test Guideline 437), the isolated chicken eye test (OECD Test Guideline 438), and reconstructed human cornea-like epithelium models (OECD Test Guideline 491) identify serious eye damage, while the short time exposure assay (OECD Test Guideline 492) and the fluorescein leakage test (OECD Test Guideline 460) support classification of mild irritants. The HET-CAM (Hen's Egg Test–Chorioallantoic Membrane) method is accepted under the Chinese Safety and Technical Standards for Cosmetics.
Skin Sensitization
Sensitization assessment is built on the adverse outcome pathway key events. The direct peptide reactivity assay (OECD Test Guideline 442C) addresses molecular initiation, ARE-Nrf2 luciferase keratinocyte reporter assays (OECD Test Guideline 442D) address keratinocyte activation, and the human cell line activation test measuring CD86 and CD54 expression (OECD Test Guideline 442E) addresses dendritic cell activation. The local lymph node assay (OECD Test Guideline 429) and the guinea pig maximization and Buehler methods (OECD Test Guideline 406) remain available where in vivo confirmation is justified. Defined approaches combining two or three of these methods are recognized by the OECD for hazard classification.
Phototoxicity
Substances absorbing UV or visible light, such as UV filters, certain botanical extracts, and some fragrance materials, require phototoxicity evaluation. The in vitro 3T3 neutral red uptake phototoxicity test (OECD Test Guideline 432) compares cytotoxicity with and without UVA irradiation and is the primary screening tool. In vivo photoirritation and photoallergy studies may be triggered by positive or equivocal in vitro results.
Genotoxicity and Mutagenicity
A standard genotoxicity battery for cosmetic ingredients comprises the bacterial reverse mutation test (OECD Test Guideline 471), an in vitro micronucleus test (OECD Test Guideline 487), and where warranted a mammalian chromosomal aberration test (OECD Test Guideline 473) or gene mutation assay (OECD Test Guideline 490). Positive findings require follow-up testing and are a significant barrier to ingredient approval, making this battery a critical component of new cosmetic ingredient dossiers.
Repeated-Dose Systemic Toxicity
Ingredients with meaningful systemic exposure potential are evaluated in 28-day (OECD Test Guideline 407) or 90-day (OECD Test Guideline 408) repeated-dose studies, typically by the oral route to provide conservative systemic hazard data. These studies establish the no-observed-adverse-effect level (NOAEL), which is essential for subsequent margin-of-safety calculations.
Reproductive and Developmental Toxicity
Reproduction and developmental toxicity screening studies (OECD Test Guidelines 421 and 422), prenatal developmental toxicity studies (OECD Test Guideline 414), and extended one- or two-generation studies (OECD Test Guideline 416) are required for ingredients with structural alerts for endocrine activity, high consumer exposure, or intended use by sensitive populations such as infants and pregnant women.
Dermal Absorption and Toxicokinetics
In vitro dermal absorption using excised human or pig skin (OECD Test Guideline 428) quantifies the systemically available fraction of an ingredient. These data feed directly into the calculation of the systemic exposure dosage and the margin of safety; where measured data are absent, conservative default absorption assumptions must be applied.
Human Predictive Testing
Human repeat insult patch tests (HRIPT), 48-hour occlusive patch tests, and photo-patch studies are conducted under ethics committee approval to confirm the cutaneous compatibility of finished formulas. These volunteer studies supplement, but do not replace, regulatory toxicology data, and they provide strong support for dermatological compatibility claims.
Testing Methods
In Vivo Legacy Methods
Historical in vivo protocols, including the rabbit Draize-type irritation studies and guinea pig sensitization assays, remain in the regulatory lexicon but are increasingly restricted. Where in vivo testing is still legally required, studies must be conducted under the OECD Principles of Good Laboratory Practice and in strict adherence to the 3Rs principles of replacement, reduction, and refinement.
In Vitro and In Chemico Alternatives
- Reconstructed human epidermis models for skin corrosion (OECD TG 431) and skin irritation (OECD TG 439)
- Transcutaneous electrical resistance assay for skin corrosion (OECD TG 430)
- Bovine corneal opacity and permeability assay (OECD TG 437) and isolated chicken eye test (OECD TG 438) for serious eye damage
- Reconstructed human cornea-like epithelium models (OECD TG 491) and short time exposure assay (OECD TG 492) for eye irritation
- 3T3 neutral red uptake phototoxicity test (OECD TG 432)
- Direct peptide reactivity assay (OECD TG 442C), ARE-Nrf2 reporter assays (OECD TG 442D), and cell line activation tests (OECD TG 442E) for sensitization
- Bacterial reverse mutation test (OECD TG 471), in vitro micronucleus test (OECD TG 487), and chromosomal aberration test (OECD TG 473) for genotoxicity
- In vitro dermal absorption through excised skin membranes (OECD TG 428)
Defined Approaches and Integrated Strategies
Defined approaches combine data from multiple non-animal methods using fixed prediction models, such as the two-out-of-three strategy for skin sensitization endorsed in OECD guidance documents. Next-generation risk assessment builds exposure-led, hypothesis-driven safety evaluations on in silico, in chemico, and in vitro data, allowing safety conclusions for many ingredients to be reached without generating any new animal data.
Standards and Guideline References
- OECD Test Guidelines for the Chemicals, maintained as the primary international method basis
- Safety and Technical Standards for Cosmetics (2015 Edition) and related GB national method standards for the Chinese market
- ISO 10993 series, including ISO 10993-10 for sensitization and ISO 10993-23 for irritation, applicable to adjacent skin-contact product categories
- ISO/IEC 17025 accreditation demonstrating laboratory technical competence
- OECD Principles of Good Laboratory Practice governing the integrity and traceability of regulatory study data
Applications
Toxicology test data generated at accredited third-party laboratories serve multiple downstream purposes across the product lifecycle.
- Regulatory dossiers supporting EU product information files and cosmetic product safety reports
- Registration and filing of imported cosmetics and special-category cosmetics under China's CSAR framework
- Safety assessment reports required for new cosmetic ingredient filings and registrations
- GHS and CLP classification of ingredients and mixtures for labeling and safety data sheets
- Margin-of-safety calculations combining NOAEL values with systemic exposure estimates
- Substantiation of compatibility claims such as hypoallergenic, dermatologically tested, and suitable for sensitive skin
- Occupational exposure assessment and manufacturing workplace safety programs
- Post-market surveillance, consumer complaint investigation, and competitive benchmarking
- Formulation research, including preservative, fragrance, and active-ingredient screening
FAQ
Is animal testing still required for cosmetics?
The European Union has banned animal testing for cosmetics since 2013. China removed mandatory pre-market animal testing for most general imported cosmetics in 2021, although certain special cosmetics, specific ingredient circumstances, and new cosmetic ingredients may still require in vivo data or safety assessment through designated channels. Many regulators now accept validated in vitro alternatives as the primary evidence base.
Which toxicology studies are typically needed for a new cosmetic ingredient?
A typical new ingredient battery includes acute toxicity, skin and eye irritation, skin sensitization, phototoxicity where the substance absorbs UV light, a genotoxicity battery, a 90-day repeated-dose study, in vitro dermal absorption, and developmental or reproductive screening triggered by exposure level and structural alerts. The exact scope depends on the target market and anticipated use concentration.
Can finished products be assessed without new testing?
Yes, under a weight-of-evidence approach. Where all ingredients have established safety profiles, the formulation review, quality control data, and published literature may be sufficient, particularly when ingredient concentrations remain below established safe limits. Formulation-specific testing is warranted when novel ingredient combinations, high active concentrations, or reactive chemistry such as oxidative hair dye systems are involved.
What is the difference between hazard identification and risk characterization?
Hazard identification determines whether a substance is intrinsically capable of causing harm, while risk characterization combines that hazard data with actual exposure conditions. The margin of safety is calculated by dividing the no-observed-adverse-effect level by the estimated systemic exposure, and a margin of at least one hundred is conventionally considered acceptable for cosmetics.
How long do cosmetic toxicology studies take?
In vitro irritation and corrosion assays are usually completed within two to six weeks. Genotoxicity batteries typically require eight to twelve weeks. Repeated-dose studies range from approximately four to six months for a 28-day study to nine to twelve months for a 90-day study including histopathology and reporting. Reproductive studies may extend beyond one year.
How should a testing laboratory be selected?
Prioritize laboratories with OECD Good Laboratory Practice compliance, ISO/IEC 17025 accreditation covering the relevant methods, validated and guideline-aligned protocols, documented experience with cosmetic matrices, and the capacity to deliver the complete study package with auditable raw data and interpretive reporting.