The practical answer
ICH Q2(R2), adopted on 1 November 2023, defines how analytical procedures are shown to be fit for purpose: specificity, accuracy, precision, detection and quantitation limits, range and response such as linearity. A certificate that says 'validated method' tells you little unless the laboratory can say which characteristics were tested, over what range and for which peptide. If the certificate says nothing, assume no validation evidence. [1] [2]
Evidence, method & source dates ↓Why validation terms matter for a purity number
A purity figure such as 98.6 percent is the output of a method. Whether it means anything depends on whether the method can separate the peptide from its impurities, how close it gets to the true value, how much it varies between runs, and how small an impurity it can see. Those are exactly the questions method validation answers.
The reference text is ICH Q2(R2), Validation of Analytical Procedures, adopted on 1 November 2023. It describes the evidence a regulatory submission should contain to show that a procedure is suitable for its intended purpose. [1] Research suppliers are not bound by it, but its vocabulary is what appears on certificates.
Glossary
Definitions paraphrase the ICH Q2(R2) glossary. [1]
| Term | Plain definition | Why it matters for a purity number |
|---|---|---|
| Specificity / selectivity | How far other substances interfere with measuring the analyte; specificity is the ideal of measuring it unequivocally | Without it, an impurity hiding under the main peak inflates purity |
| Accuracy | Closeness of the measured value to an accepted reference value | A precise method can still be consistently wrong |
| Precision | Scatter between measurements of the same homogeneous sample, usually given as standard deviation or coefficient of variation | Tells you whether 98.6 vs 98.1 is a real difference |
| Repeatability | Precision under the same conditions over a short interval | Best case variation, same analyst and run |
| Intermediate precision | Variation within one laboratory across days, analysts or equipment | Closer to what you see lot to lot |
| Reproducibility | Precision between laboratories | Why two labs can report different purities |
| Detection limit (DL or LOD) | Lowest amount that can be detected but not necessarily quantified | Impurities below it are invisible |
| Quantitation limit (QL or LOQ) | Lowest amount that can be quantified with suitable precision and accuracy | Impurities below it cannot be reported as numbers |
| Response and linearity | Signal relates to amount by a known function; linearity is one such model | Matters whenever a result is calculated from a calibration |
| Reportable range | Lowest to highest result with suitable precision and accuracy | A result outside it is not supported |
| Working range | Lowest to highest level as presented to the instrument with reliable results | Sample preparation must land inside it |
| Robustness | Ability to meet performance criteria under deliberate small changes to method parameters | Explains why a small change in column or gradient can shift results |
| System suitability test | Routine check that the system is fit for purpose on the day | Ask whether it passed for your run |
| Platform procedure | One procedure used for sufficiently similar molecules without significant changes | A generic peptide method may not suit every sequence |
| Revalidation / co-validation | Showing fitness again after a change, or across laboratories | A method validated elsewhere still needs evidence at this laboratory |
Which characteristics apply to which test
Q2(R2) does not ask for every characteristic for every test. Its Table 1 sets out the typical pattern: identity tests need specificity; quantitative impurity tests need specificity, response, a quantitation limit, accuracy, repeatability and intermediate precision; impurity limit tests need specificity and a detection limit; assays for content or potency need specificity, response, accuracy and precision. [1] A peptide purity result by HPLC is effectively an impurity measurement, so the quantitation limit matters.
| Characteristic | Identity | Impurity, quantitative | Impurity, limit test | Assay (content) |
|---|---|---|---|---|
| Specificity | Yes | Yes | Yes | Yes |
| Response (calibration model) | No | Yes | No | Yes |
| Lower range limit | No | QL | DL | No |
| Accuracy | No | Yes | No | Yes |
| Repeatability | No | Yes | No | Yes |
| Intermediate precision | No | Yes | No | Yes |
The numbers the guideline uses
These figures let you judge whether a laboratory's validation summary is in the expected territory. All are from ICH Q2(R2). [1]
- Detection limit by signal-to-noise: a ratio of 3:1 is generally acceptable.
- Quantitation limit by signal-to-noise: a ratio of at least 10:1 is acceptable.
- From a calibration line: DL = 3.3 sigma / S and QL = 10 sigma / S, where sigma is the standard deviation of the response and S the slope.
- For impurity tests, the quantitation limit should be at or below the reporting threshold.
- Linearity during validation: at least five concentrations.
- Repeatability: at least 9 determinations across the range (for example 3 concentrations, 3 replicates each) or at least 6 at 100 percent of the test concentration.
- Reportable range for a product assay: typically 80 to 120 percent of the declared content; for impurities, from the reporting threshold to 120 percent of the specification.
'Validated', 'qualified', 'verified' or nothing
Validated should mean a documented study against predefined criteria, summarised in a report. Q2(R2) says a validation protocol should be written before the study and results summarised in a validation report. [1] ICH Q7 says analytical methods should be validated unless they are pharmacopoeial or another recognised standard method, and that even those should be verified as suitable under actual conditions of use. It also says the degree of validation should reflect the purpose of the analysis, and that modifications to a validated method need records showing the change gives equally accurate and reliable results. [2]
Qualified is ambiguous on a certificate. In ICH Q7 qualification is applied to equipment, and the guide says appropriate qualification of analytical equipment should be considered before method validation starts. [2] A 'qualified method' may mean anything from a full validation to a single check; ask.
Silence is common. If a certificate gives a purity number with no method details and no validation statement, treat it as an unvalidated result. Regulated manufacturers relying on a supplier's certificate must still run an identity test and validate the supplier's results at intervals, which shows how much weight regulators put on an unsupported certificate. [3]
Questions to ask the testing laboratory
Bio-Synthesis describes its HPLC deliverable as an annotated chromatogram with a purity value and a method summary, with the peak table on request. [4] That method summary is where these answers should start.
- Is the method validated, verified or neither? Can you share the validation summary?
- Was specificity shown for this peptide or a similar one, and how, for example by an orthogonal method?
- What is the quantitation limit for impurities, and what reporting threshold did you use?
- What precision do you see for this method, and is that repeatability or intermediate precision?
- Did system suitability pass for my run?
- Is this a platform method used for many peptides? Was anything adjusted for this sequence?
- Can I have the peak table and the raw data file as well as the PDF?
Questions before you decide
What is the difference between LOD and LOQ on a certificate?
Under ICH Q2(R2), the detection limit is the lowest amount that can be detected, while the quantitation limit is the lowest amount that can be measured with suitable precision and accuracy. Impurities between them are seen but not quantified.
Does 'validated method' on a COA mean the purity is reliable?
Not by itself. Ask which characteristics were validated, for which peptide and over what range. ICH Q2(R2) expects a protocol and a validation report, so a laboratory should be able to summarise them.
What signal-to-noise ratio does ICH Q2(R2) accept for detection and quantitation limits?
About 3:1 is generally acceptable for estimating the detection limit, and at least 10:1 for the quantitation limit.
What does it mean if a certificate says nothing about method validation?
Treat the result as unsupported by validation evidence. Even regulated manufacturers relying on supplier certificates must test identity and periodically check the supplier's results.
Check the underlying evidence
Sources & review dates
- ICH Q2(R2) Validation of Analytical Procedures, final version adopted 1 November 2023 ↗Accessed 30 Oct 2026
- ICH Q7 Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients ↗Accessed 30 Oct 2026
- 21 CFR 211.84 Testing and approval or rejection of components (Cornell LII) ↗Accessed 30 Oct 2026
- Bio-Synthesis peptide analytical services: sample amounts, information required and deliverables ↗Accessed 30 Oct 2026
Sources are documentary evidence. They do not establish firsthand testing of a listed supplier.