A COA records tests on one batch at one point in time. A stability study tracks change over time, so the two documents answer different questions.

A COA records tests on a batch. A stability study tracks measured changes over time under stated storage conditions.

A certificate of analysis (COA) reports tests run on a named batch of material. It is a snapshot of test results from a particular time, not a record of how the material changes later.
A useful COA pairs each measured result with its test method and specification. The specification is the allowed range or limit, while the result is what the test found.
For example, a test might report identity, purity, or content. Content means the measured amount of the intended material; purity describes the share of the tested sample that consists of that material rather than detected impurities.
High-performance liquid chromatography (HPLC), which separates parts of a mixture as it passes through a column, can support a purity estimate. Mass spectrometry measures the mass of charged molecules and can support identity, but it does not by itself measure total content or every impurity.
Those method details define what a COA can support. Results for one tested attribute do not prove other attributes, and a COA cannot establish long-term shelf life on its own.
Our explanation of how to read a certificate of analysis covers the fields and terms that help make test results clear.
Start with the lot or batch number on the COA and compare it with the material received. A mismatch breaks the link between the document and that material, so the results cannot clearly describe the batch in hand.

Traceability means a clear record linking the sample, batch, and test. A useful COA makes that link easy to follow and identifies the material tested.
Check the test date, method, measured result, and pass range for each reported test. Together, these details show what the lab tested, when it tested it, and what limits it used.
Look for a named lab and enough detail to identify the test method. If a test result says only pass, without a method or measured value, it gives less detail for judging the evidence.
A generic COA with no clear batch link is weak evidence, even if it lists test names or pass ranges. A lab report also describes the sample tested, so sampling and handling records help show how that sample relates to the batch.
For more on test records and lab methods, see our lab testing information. Read each result beside its method and specification, rather than treating a list of test names as proof of every quality claim.
A stability study tests samples at planned time points under stated storage conditions. It checks whether selected qualities change as time passes.
A clear report sets out the study plan, the batch or batches tested, storage conditions, time points, test methods, results, and acceptance criteria. Acceptance criteria are the limits that results must meet.
The report should show results from the start and later time points. Comparing those results reveals whether measured qualities stayed within the stated limits or moved toward them.
Read the trend, not only the pass-or-fail result. A measurement can stay within its acceptance range while moving consistently toward a limit; that pattern deserves attention, even though it is not itself a failure. Compare like-for-like results from the same method and conditions, and consider whether a change is larger than expected test variation before attributing it to the material.
Methods should be comparable across time points. If the method changes, a difference in results may come from the test rather than a change in the material.
Studies may track identity, content, purity, appearance, or other qualities that matter for that material. The report needs to name the attributes tested, because results cannot support claims about qualities the study did not measure.
Accelerated testing uses more stressful storage conditions, such as higher heat, to make some changes appear sooner. It can add useful evidence, but its results do not automatically establish how long material will remain within limits under long-term storage.
Set time points in advance: establish a baseline, test at planned intervals under the intended long-term storage conditions, and include a final point that reaches the period being claimed. Keep accelerated samples in a separately identified arm; they can help reveal changes sooner but do not replace long-term observations. There is no single study duration for every material: long-term plans are generally built around the intended shelf-life claim, and a changed claim calls for checking whether the available time points still support it.
Stability results apply to the material and conditions studied. A change in the formula, production process, packaging, or storage setting can limit how far those results carry over.

Check whether the study conditions match the intended storage conditions. Results from one temperature or package may not support a claim for a different setting.
Packaging matters because it can affect how much light, air, or moisture reaches the material. A report should identify the package used, especially when the proposed storage claim depends on that package.
Storage during shipping can also matter when it differs from the study conditions. Our article on temperature changes during transit explains why transport records can help put storage evidence in context.
A study of one batch does not show that every later batch will behave the same way. Testing more batches over time can add evidence about how consistently the material meets its limits.
A date printed on a COA does not, by itself, prove shelf life. Look for stability results that cover the claimed period and storage conditions.
A retest date marks when material should be tested again to see whether it still meets its limits. An expiration date marks the end of its stated usable period.
These dates answer different questions. A retest date calls for a new check, while an expiration date sets the end of the stated period.
Compare a proposed date with the study's time points, results, and pass ranges. If the study ends before the claimed period, its results do not directly show what happened after the last tested time point.
A short statement that material is stable is not a substitute for the study report. The report should show the methods and results behind the claim, along with the conditions and time points used.
Use a COA to see whether a tested batch met its stated limits at the time of testing. Use stability data to see how selected qualities changed over time.

Batch release testing and ongoing stability monitoring can support each other. The first checks a batch at a point in time; the second tracks selected qualities across a planned period.
| Document | Main question | Evidence shown | Key limit |
|---|---|---|---|
| COA | Did this batch meet its limits? | Test results at a stated time | Does not show change over time |
| Stability study | Did selected qualities change? | Results across planned time points | Applies to tested material and conditions |
| COA and study together | How did a tested batch perform over time? | Batch tests and stability trends | Claims remain limited to the evidence |
If a COA makes a stability claim, ask for the study details behind it. The report should identify the batches, storage conditions, time points, methods, results, and acceptance criteria.
A COA and a stability report can refer to related material, but they are not interchangeable. A batch-specific COA cannot replace a record of change over time, and a stability report does not automatically prove that every batch passed its release tests.
The two documents answer different questions. A COA records batch results at test time; a stability study tracks change under stated conditions.
Read them together, and match every conclusion to its batch, method, limits, and storage setting.