Lab Testing & COA
How to Read a Bacteriostatic Water Certificate of Analysis
Most bacteriostatic water listings say ‘lab tested’. Far fewer publish the document. Fewer still publish one you can match to the vial in your hand and verify with the laboratory that signed it.
A certificate of analysis records what one specific production lot measured, on a specific date, by a named method. Here is how to read one — using the certificate for the lot shipping from us right now as the worked example.
Published by BacterioWater7 min read

In short
- A certificate describes one lot. A specification sheet describes an intention. They are not the same document.
- The methods column is the part worth reading: it tells you whether a result was measured or asserted.
- For a preserved water, the certificate should both identify and quantify the preservative.
- The lot code on the vial must match the lot code on the certificate, or the certificate is about someone else’s water.
- A certificate number and access code you can check with the issuing laboratory is what turns a PDF into evidence.
What a certificate of analysis is — and is not
A certificate of analysis (COA) is a record issued by the testing laboratory for one production lot. It lists each test performed, the specification that test was judged against, the result obtained, and a conformance call. Everything on it is retrospective: it describes material that already exists.
Three documents it is regularly confused with
- A specification sheet
- States what a product is intended to be. It is written before manufacture and applies to every lot in the abstract, which means it can be true and still tell you nothing about your vial.
- A safety data sheet
- Covers hazard classification, handling and first aid for a substance. Useful, required, and entirely unrelated to whether a given lot met its specification.
- A certificate of conformance
- A supplier’s own statement that a lot conforms. It carries the supplier’s word rather than a laboratory’s measurements, and it usually has no results table at all.
None of these is a substitute for a per-lot certificate, and ‘COA available on request’ is not the same as a published one. Ours are published — the current lot’s certificate sits on the Lab Testing & COA page as an image of the issued document, not a retyped summary.
The header block
Before the results, a certificate identifies what was tested and who tested it. Every field in that block does work:
- Product and batch code
- The lot the certificate covers. This is the field you match against the code printed on the vial label — on the certificate worked through here, that is batch EB61901.
- Laboratory
- Who performed the analysis. A named independent laboratory is a different assurance from an in-house QC line, and the certificate should make clear which one you are looking at. Ours was issued by Chromate.
- Certificate number
- The laboratory’s own reference for the document — the key you use to look it up with them. Ours is #35348.
- Sample received, analysis conducted, issued
- Three separate dates. On lot EB61901: received 12 May 2026, analysed 22 May 2026, issued 24 May 2026. The gaps are what a real testing schedule looks like.
- Appearance
- A recorded observation of the sample as received — ‘clear liquid’ for this lot. Small, but it is an observation rather than a claim.
- Signatory
- A named analyst who takes responsibility for the document. Lot EB61901 is signed by Lucas Weber, Principal Chemist.
The results table, line by line
A results table has four columns for a reason: a test, the specification it was judged against, the measured result, and the conformance call. Read the middle two together — ‘Conforms’ on its own is a comparison, not a grade.
| Row | What it establishes | What a weak certificate does instead |
|---|---|---|
| Identity | Confirms the preservative present really is benzyl alcohol. | Names the preservative in the product description and never tests for it. |
| Quantity | Measures how much of it is there, against the formulation target. | Reports ‘0.9%’ as a specification with no measured value beside it. |
| Sterility | Confirms no viable microbial growth is recovered from the finished vial. | Says ‘sterile filtered’ — a process step, not a result. |
| Endotoxins | Quantifies bacterial endotoxin load against a stated limit. | Reports ‘pass’ with no number and no limit. |
| pH | A measured value inside a stated range. | States a range only, as though it were a result. |
For comparison, this is the full results table as issued for the lot shipping now:
| Test | Specification | Result | Conformance |
|---|---|---|---|
| Identity | Benzyl alcohol | Benzyl alcohol | Conforms |
| Quantity | 0.9% | 0.921% | Conforms |
| Sterility | Pass | Pass | Conforms |
| Endotoxins | < 0.5 EU/ml | < 0.25 EU/ml | Conforms |
| pH | 4.5 – 7.0 | 6.253 | Conforms |
Note what the numbers do. Quantity, endotoxins and pH all return measured values rather than a bare pass, which means a reader can see how much margin the lot had against its specification instead of taking the conformance call on trust.
Read the methods, not just the conformance column
Two certificates can both say ‘Conforms’ on every row and be worth very different amounts. The methods statement is where that difference lives.
Identity and quantity — RP-HPLC with UV detection
Reversed-phase HPLC separates the components of the solution and a UV detector quantifies them. The evidence is a chromatogram with a resolved peak at a known retention time — 4.358 minutes for benzyl alcohol on this lot. A certificate that reports a preservative concentration without naming a separation method is reporting a formulation target.
Sterility — the compendial test and its alternatives
USP <71> is the compendial sterility test. Automated growth-detection systems such as BacT/ALERT are alternative microbiological procedures, validated against the compendial method under USP <1223>. Either is defensible; what matters is that the certificate says which one was used. Lot EB61901 states sterility by USP <1223> BacT/ALERT.
Endotoxins — a number, not a verdict
USP <85> permits gel-clot, turbidimetric and chromogenic techniques. Gel-clot returns a pass/fail against a threshold; kinetic chromogenic LAL is quantitative and returns a value. This lot reports < 0.25 EU/ml against a < 0.5 EU/ml specification, which is a stronger statement than ‘pass’ on the same row.
pH
The least glamorous row and a genuinely useful one: a measured pH sitting mid-range (6.253 against a 4.5–7.0 specification) is consistent with a solution that was made to formulation and has not drifted.
Verify it at the source
A published PDF proves that a document exists. Verification proves the laboratory issued it. The two are not the same and only one survives scrutiny.
- Find the certificate number and access code on the document. On lot EB61901 those are certificate #35348 and access code EBAC444D2MUF.
- Enter them with the issuing laboratory — for this certificate, at chromate.org/verify.
- Match the batch code on the certificate to the code printed on the vial label you received. If they differ, the certificate is about a different production run.
A seven-point check before you buy
- Is there a certificate published, rather than ‘available on request’?
- Does it name a lot, and can you match that lot to the vial you would actually receive?
- Was it issued by a named laboratory that is independent of the seller?
- Is the preservative both identified and quantified, with the method named?
- Is sterility reported as a result, with the test method stated?
- Is the endotoxin result a number against a stated limit, rather than ‘pass’?
- Can the certificate be verified with the issuing laboratory using a reference and access code?
Seven out of seven is not a high bar, but it is not commonly met. Ours is published lot by lot in the COA library, and the 30 mL vial that ships carries the matching code on its label.
For research use only. Not for human or veterinary use. Nothing in this article is a dosing, protocol, or medical instruction. Read the full RUO disclaimer.
References
USP <71> Sterility Tests
The compendial sterility test method.
USP <1223> Validation of Alternative Microbiological Procedures
How automated growth-detection systems are validated against the compendial test.
USP <85> Bacterial Endotoxins Test
Gel-clot, turbidimetric and chromogenic techniques, and how results are expressed.
Certificate of Analysis #35348, lot EB61901 — Chromate
The worked example throughout this post, published in full.
Questions this raises
Does every lot need its own certificate?
Yes, if the certificate is doing any work. A certificate covers the lot that was sampled and tested. Reusing an older certificate for newer production tells you nothing about the vial in the box.
What is the difference between a COA and a specification sheet?
A specification sheet states the target a product is made to. A certificate of analysis states what was actually measured on one lot. A supplier who publishes only the first has published an intention, not a result.
What does ‘Conforms’ mean on a certificate?
That the measured result for that row fell inside the specification listed beside it. It is a comparison against a stated limit, not an independent grade, so the specification and the result are the two values worth reading together.


