Information requirements and test methods and quality of data

According to which test methods and standards should new tests be performed?

When you perform new tests you have to follow the Test Method Regulation (Commission Regulation No 440/2008) or another method recognised by the European Commission or ECHA (Article 13(3) of REACH).

In REACH Annexes VII to X on standard information requirements, the use of various OECD test guidelines is required (e.g. OECD TG 414, 421 and 422).

The OECD methods can be found at: http://www.oecd-ilibrary.org/

The text of the Test Method Regulation can be found at: http://eur-lex.europa.eu/JOHtml.do?uri=OJ:L:2008:142:SOM:EN:HTML

Note that changes occurred in the Annexes VII and VIII requirements in 2016 for the following endpoints: skin and eye irritation, skin sensitisation and acute dermal toxicity, making non-animal testing the default.

Information may be generated using other methods (Article 13(3)) provided the conditions defined in Annex XI are met. These include amongst others that the result is sufficient for the purposes of classification and labelling and/or risk assessment, and that adequate and reliable documentation of the applied method is provided (see Annex XI for more information).

Furthermore, a specific requirement exists for ecotoxicological and toxicological tests: new tests have to be carried out in compliance with the principles of good laboratory practice (GLP) provided for in Directive 2004/10/EC, as no other international standard has so far been recognised as being equivalent. For physicochemical testing, it may be desirable but it is not mandatory to have tests performed according to the GLP standard.

The Guidance on Information Requirements and Chemical Safety Assessment contains specific integrated testing strategies for each endpoint (e.g. for aquatic toxicity, mutagenicity), which should be consulted before new tests are performed. You can find this document at: http://echa.europa.eu/guidance-documents/guidance-on-information-requirements-and-chemical-safety-assessment

Are there "other international test methods" recognised by the Commission or the ECHA and referred to in article 13(3) of REACH?
For the time being, no "other international test methods" within the meaning of Article 13(3) of the REACH Regulation have been recognised by the Commission or by ECHA.
Is there a list of GLP-certified testing laboratories?

The European Commission has published the lists of inspected test facilities provided by national GLP monitoring authorities. You can access the document at:

Good laboratory practice (GLP) certification of laboratories is the responsibility of national authorities that administer the national monitoring programmes.

If the laboratory is located in the EU, Norway or Switzerland, the corresponding authority can be found at the website of DG Enterprise and Industry of the European Commission at:

If the laboratory is located in another country, you should check the section on GLP of the OECD website:

After you have identified the relevant GLP monitoring authority, you can consult this authority to find out the laboratories with GLP certification in the corresponding country.

Additionally, laboratories can be inspected by a GLP monitoring authority even if located in a country which has not joined the OECD mutual acceptance of data system. Information on these laboratories can be obtained from the GLP monitoring authority who has inspected them (see also Q&A 122 ).

Are reference books and databases regarded as reliable sources of substance data?

In general, there is the possibility to use data from reliable, scientifically accepted reference literature or databases, provided that the substance to be registered and the substance described in the reference are comparable with regard to homogeneity, impurities, particle size etc.

The documentation of similarity needs to be submitted in the registration dossier. References to literature or databases often use secondary data sources. When such data is used, the original source should be cited and checked by an expert.

Some useful reference books and data compilations containing peer reviewed data are listed under each endpoint in the Guidance on Information Requirements and Chemical Safety Assessment, Chapters R.7 a, b, c: Endpoint specific guidance available on the ECHA website at: http://echa.europa.eu/guidance-documents/guidance-on-information-requirements-and-chemical-safety-assessment.

For some endpoints, these data from reference literature or databases may be used on their own to fulfil the information requirement. However, in general they will have to be combined to other pieces of evidence and submitted as part of a weight-of-evidence approach or read-across approach to support the justification proposed to adapt the requirement.

What is the OECD mutual acceptance of data (MAD) system?

The OECD decision on mutual acceptance of data (MAD) provides for data generated by testing of chemicals in an OECD member country in accordance with OECD test guidelines and OECD principles of good laboratory practice (GLP) to be accepted in other member countries for purposes related to the protection of human health and the environment.

This system also covers non-OECD countries, which have requested adherence to the OECD GLP and to join the MAD system. These non-OECD countries can be divided into two groups:

  1. Countries that are full adherents to the OECD MAD system.
  2. Countries that are provisional adherents to the OECD MAD system.

Countries that are full adherents to the OECD MAD system will accept data from OECD member countries and other adhering countries generated under MAD conditions. In addition, non-clinical safety data developed in these countries must be accepted by the OECD and adhering countries.

Countries that are provisional adherents to the OECD MAD system need to accept data from OECD member countries and other adhering countries generated under MAD conditions. However, during the period of provisional adherence, GLP monitoring activities conducted by the GLP monitoring authority located in the country of the provisional adherence do not have to be accepted by the full members of the OECD MAD decision.

What studies does ECHA accept as GLP studies?

In general, ECHA accepts data as GLP data where this data comes from a test facility:

  • from countries that are OECD member states or full adherents to the OECD mutual acceptance of data (MAD) system; and
  • from countries that are provisional adherents to the OECD MAD system and in which laboratories have been inspected jointly by the GLP monitoring authority concerned and by an OECD GLP monitoring authority.

Studies that are conducted in a test facility situated in a country which has not joined the OECD MAD system can be accepted by ECHA as GLP compliant studies under the following conditions:

  • Before performing the study, the GLP compliance of the test facility has been inspected by: 
    • an EU GLP monitoring authority (including Norway through EEA agreement); or GLP monitoring authorities in Israel, Japan and Switzerland with whom the EU holds mutual recognition agreements; or 
    • other GLP monitoring authorities of OECD member states or full adherents to the OECD MAD system on a case-by-case basis; or
    • other national GLP monitoring authority which has been assessed on-site by representatives of the EU GLP Working Group and whose Compliance Monitoring Programme could be regarded as being equivalent to the EU GLP Compliance Monitoring Programme; and
  •  The test facility has been found to be operating in compliance with GLP principles.


See also Q&A 119.

Registrants who submit a proposal for testing in accordance with Annexes IX and X of REACH may waive 28-day studies if certain conditions are fulfilled. However, if there are no results for a 28-day repeated dose toxicity study because a testing proposal for a 90-day repeated dose toxicity test is made, it is not possible to derive a DNEL. Which interim risk management measures (RMM) could be recommended in this situation?
Interim RMM are to be included into the Chemical Safety Report (CSR) and to be communicated to the users of the substance (either under Article 31 or 32 of REACH, in form of Exposure Scenario (ES) or otherwise). Such RMMs need to describe how to handle a substance in an appropriate way in the absence of sufficient toxicological information and while waiting for the results of proposed testing for long-term hazards. If no Derived No-Effect Level (DNEL) is available, the registrant is expected to carry out a qualitative risk characterisation referring to the identified uses and the expected exposure and justifying that the measures are sufficient to control the risks (based on the available knowledge). Typical RMM applicable in such case are listed in the Table E.3-1 of Part E: Risk Characterisation of the Guidance on information requirements and chemical safety assessment available at: http://echa.europa.eu/guidance-documents/guidance-on-information-requirements-and-chemical-safety-assessment
Where no or not enough hazard information on the substance is available, the registrant should demonstrate control of risks by minimizing the emission and/or exposure to the substance. In doing so, he can use a combination of containment and/or Local Exhaust Ventilation (LEV) and/or Personal Protective Equipment (PPE) as interim RMM to protect workers from exposure. Due to the interim nature of the measures, PPE may play a more prominent role compared to what is suggested in table E.3-1 of the abovementioned Guidance.
The registrant may need to update his CSR and ES once he receives the result of the test proposal and is able to derive a DNEL (or identify that minimisation of emission/exposure is required if "no threshold" effects had been identified in the testing). This applies in particular if the interim measures had been based on PPE instead of containment or other engineering measures.
When a substance is classified only as an eye irritant, does a full risk characterisation as part of the chemical safety assessment (CSA) need to be carried out?

Not necessarily. The scope of the risk characterisation, that you have to carry out as part of the CSA, depends on the hazard profile of the substance. It has to address every hazard, not just those that lead to a classification (points 0.5 and 6 in Annex I to REACH).

Firstly, you have to consider each physical, health and environmental hazard identified, even if classification is not required. This includes collecting the predicted or derived no-effect levels or minimal effect levels (PNECs, DNELs or DMELs) if appropriate.

You should also consider the relevant timescales, environmental compartments, human populations, health effects, and routes of exposure.

DNELs for irritation/corrosion can only be derived if dose-response information is available. Therefore, for endpoints such as eye irritation where no DNEL can be derived, a more qualitative approach to assessing and controlling such risks is necessary. This may be the case where the pH led to the classification or where only QSAR data are available.

For information about this approach, see Chapters R.8 (Part E) and R.10 of the Guidance on Information Requirements and Chemical Safety Assessment: http://echa.europa.eu/guidance-documents/guidance-on-information-requirements-and-chemical-safety-assessment.

If there are no other hazards, then it is sufficient to describe the measures which ensure that the risks to eyes are avoided or managed in the exposure scenarios (ESs). If there are other hazards identified, then your assessment should address these also.

The exposure assessment and the subsequent risk characterisation should cover all stages of the life cycle of the substance resulting from the substance's manufacture, and the identified uses.

The Practical Guide on How to undertake a qualitative human health assessment and document it in a chemical safety report is a helpful document when undertaking a qualitative human health assessment: http://echa.europa.eu/practical-guides

Tip: Using Chesar will help you determine the scope of exposure assessment and the type of risk characterisation. For more details on this see the ‘Chesar User Manual, Part 1, Section 6' available at: http://chesar.echa.europa.eu/web/chesar/support/manuals-tutorials

What are the criteria for deciding if a substance is highly insoluble in water or poorly water soluble?

For aquatic toxicity testing (Section 9.1 Annexes VII and VIII), Column 2 adaptations include two complementary concepts related to solubility in water.

The concept of "highly insoluble in water" is associated with the likelihood for aquatic toxicity; consequently a general threshold cannot be established. The use of this concept for waiving aquatic toxicity testing requires substance-specific assessment.

In the waiving statement, registrants should justify that aquatic toxicity is unlikely to occur at the limit of the water solubility. This may require specific information, such as that obtained from transformation/dissolution studies or from the identifying the components of the water accommodated fraction (see the webinar presentation: Hints and Tips on Physicochemical, environmental and human health related endpoints - Aquatic Toxicity).

If registrants cannot demonstrate that aquatic toxicity is unlikely to occur, the substance should be considered as "poorly water soluble", not as "highly insoluble in water", and therefore long-term testing has to be considered.

The concept of "poorly water soluble" is associated with the need to consider long-term tests instead of short-term tests. The ECHA Guidance on Information Requirements and Chemical Safety Assessment section R.7.8.5 (Endpoint Specific Guidance R.7.b) suggests that water solubility below 1mg/L or below the detection limit of the analytical method of the tested substance should be used for considering the substance as poorly water soluble and performing the long-term tests instead of the short-term tests:

For further details regarding testing on aquatic toxicity please consult OECD Guidance Document on Aquatic Toxicity Testing of Difficult Substances and Mixtures available at: