ISO 17034 vs ISO/IEC 17025: What Your Certificate of Analysis Really Tells You

Published 16 July 2026 · By Andrew Keating, Head of Quality and Accreditation

Few documents receive less attention and carry more weight than a certificate of analysis. It arrives in the box, it is filed, and it is only read closely when an assessor asks a question or a result is challenged. At that point the differences between accreditation schemes stop being administrative detail and start determining whether your data holds up.

Two standards, two different subjects

The most common misunderstanding is that ISO 17034 and ISO/IEC 17025 are alternative routes to the same assurance. They are not. They accredit different activities.

• ISO 17034 sets general requirements for the competence of reference material producers. It covers production planning, raw material control, processing, homogeneity assessment, stability studies (both long-term and under transport conditions), value assignment, uncertainty estimation, commutability where relevant, and the content of the certificate itself.
• ISO/IEC 17025 sets requirements for the competence of testing and calibration laboratories. It covers method validation, equipment calibration, personnel competence, measurement traceability and the reporting of results.
• ISO/IEC 17043, often held alongside both, covers the competence of proficiency testing providers.

A producer can hold ISO/IEC 17025 for its analytical measurements without holding ISO 17034 for material production. In that case the numbers on the certificate may be perfectly sound as measurements, while the homogeneity, stability and value assignment framework that turns a measured batch into a certified reference material has not been independently assessed.

What each certificate type is actually claiming

In practice, materials fall into a spectrum:

1. Certified reference material produced under ISO 17034. Carries a certified property value, an expanded uncertainty with coverage factor, a documented traceability statement, homogeneity and stability data, and defined storage and handling conditions. This is what regulatory methods mean when they specify a CRM.
2. Reference material with a characterised value. May carry an assigned value from a competent measurement but without full homogeneity or stability assessment, and often without an uncertainty budget.
3. Analytical standard with a purity statement. Suitable for identification or qualitative work, but the purity figure may be a chromatographic area percentage rather than a mass balance assay — a distinction that materially affects quantitative results.

None of these is inherently wrong. Using the third where a method requires the first is.

How to read a certificate properly

A certificate that will withstand scrutiny answers these questions without you having to telephone the supplier:

• What exactly is the certified value, and is it a mass fraction, a concentration, an acid or an anion basis?
• What is the expanded uncertainty, and what coverage factor and confidence level apply?
• How was the value assigned — mass balance, quantitative NMR, interlaboratory consensus, or a reference method?
• To what is the value traceable, and through what chain?
• What is the minimum sample intake below which homogeneity is not guaranteed?
• What are the storage conditions, the retest date and the stability basis for that date?
• Which accreditation body and scope number cover this material, and is production explicitly within that scope?

The last point is worth emphasising. An accreditation logo on a document does not mean every product from that organisation is within the accredited scope. Scopes are specific, published and checkable.

Where laboratories get caught at assessment

The recurring findings are predictable. A method that specifies a certified reference material but is run with a non-certified analytical standard. An uncertainty budget that omits the contribution of the reference material entirely. A purity value transcribed into the LIMS without the associated uncertainty. Expired material still in use because the retest date was recorded on the certificate but not in the inventory system. And second-source verification performed against a different lot from the same producer, which does not provide genuine independence.

A short internal checklist

1. For every method, record whether it requires a CRM or an analytical standard, and why.
2. Store certificates so they are retrievable against the batch record, not only the purchase order.
3. Include the reference material uncertainty in every quantitative uncertainty budget.
4. Verify supplier accreditation scopes annually rather than assuming continuity.
5. Use genuinely independent second-source material for calibration verification.

The underlying point

Accreditation exists so that a laboratory in one country can rely on a measurement made in another without renegotiating trust every time. ISO 17034 and ISO/IEC 17025 are two links in that chain, and they are not interchangeable. Knowing which link your certificate represents is the simplest possible safeguard for the defensibility of everything you report.