How to Read a Certificate of Analysis — A Guide for Quality Managers

The CoA Is Not a Guarantee

Every raw material shipment arrives with a certificate of analysis. Most quality teams glance at the results column, confirm the numbers fall within specification, and file the document. That approach leaves significant risk on the table.

A CoA is a supplier’s representation of a lot’s test results. It is not independently verified unless you verify it. It may reflect testing performed months before shipment. It may have been generated from a composite sample that does not represent the lot you received. In some cases — which we will address in a separate post on CoA fraud — it may not reflect actual testing at all.

Reading a CoA critically is a core quality competency. This guide walks through each section of a well-structured CoA and explains what to look for, what to question, and what should trigger escalation.

---

Section 1: Lot and Traceability Information

The first thing to verify is that the CoA matches the physical shipment. This sounds obvious, but lot number discrepancies are more common than they should be.

Check:

• Lot/batch number: Does it match the label on the container and the purchase order?
• Manufacture date and expiration/retest date: Is the material within its shelf life? If the retest date has passed, the CoA results may no longer be valid — consult your specification for retest requirements.
• Quantity: Does the CoA quantity match what was received?
• Country of origin: Relevant for regulatory compliance and risk assessment. A change in country of origin from a previously qualified supplier should trigger re-qualification review.

If any of these fields are blank, ask the supplier to provide a complete document before accepting the lot.

---

Section 2: Product Description and Specifications

Check:

• Product name and grade: Confirm it matches your purchase specification exactly. “Ashwagandha root extract” and “ashwagandha extract” are not interchangeable if your specification calls for a specific plant part.
• Specification version: Does the CoA reference a specification version? If the supplier has updated their spec without notifying you, results may be passing against a different standard than yours.
• CAS number or INCI name: For chemical ingredients, verify the CAS number matches the substance you ordered. For cosmetic ingredients, confirm the INCI name.

---

Section 3: Test Results

This is the section most quality teams focus on — and it deserves careful reading, not just a pass/fail scan.

For each test parameter, verify:

1. The result vs. the specification limit Confirm the result falls within your specification, not just the supplier’s specification. These may differ. If your spec requires ≥5% withanolides and the CoA shows 5.2% against a supplier spec of ≥4%, the material passes your spec — but only just. Flag lots that pass by a narrow margin for trend monitoring.

2. The unit of measurement “Lead: 0.5” is meaningless without units. Is it 0.5 ppm? 0.5 mg/kg? 0.5 µg/g? These are equivalent, but 0.5 mg/L is not. Confirm units are stated and match your specification.

3. The test method Every result should reference a test method — USP <561>, AOAC 2007.01, ISO 11290-1, an in-house validated method (with version number), etc. A CoA that lists results without method references cannot be evaluated for fitness-of-purpose.

4. The detection/reporting limit For trace analytes (heavy metals, pesticides, microbial counts), a result of “ND” (not detected) or “<X” is only meaningful if the reporting limit is below your specification limit. If your lead limit is 0.5 ppm and the lab’s reporting limit is 1.0 ppm, a “ND” result does not confirm compliance.

5. The instrument or technique For analytical chemistry tests, the instrument type matters. “Heavy metals: pass” with no method or instrument reference is not an acceptable result for a quality-critical raw material.

---

Section 4: Microbial Results

Microbial results require particular attention because the test methods and enumeration limits vary significantly between pharmacopeial standards.

Check:

• Which pharmacopeia was used? USP <61>/<62>, EP 2.6.12/2.6.13, and JP methods have different acceptance criteria and procedural requirements. A result that passes USP may not pass EP limits, and vice versa.
• Are all required organisms tested? For dietary supplements, USP <2023> specifies the required microbial tests by product category. For cosmetics, ISO 17516 defines limits. Confirm the CoA covers all organisms required by your specification.
• Are the results reported as CFU/g or CFU/mL? Confirm the unit matches your specification.
• What was the sample size? USP <61> requires a minimum sample size for enumeration. A result based on an inadequate sample size is not valid.

---

Section 5: Analyst and Laboratory Information

Check:

• Laboratory name and accreditation: Is the testing laboratory ISO 17025 accredited for the specific tests listed? Accreditation scope is test-specific — a lab may be accredited for microbiology but not for pesticide residue testing.
• Analyst signature and date: The CoA should be signed (or digitally authorized) by a qualified analyst or quality manager. An unsigned CoA is not a complete document.
• Testing date vs. issue date: If the testing date is significantly older than the issue date, ask why. A CoA issued months after testing may indicate the results were held pending investigation.

---

Section 6: Statements and Disclaimers

Many CoAs include statements such as “This CoA is valid for the lot number listed above” or “Results are based on representative sampling.” Read these carefully. They define the scope of the supplier’s representation and may limit your recourse if the material does not perform as expected.

---

Practical Checklist: CoA Review Before Lot Acceptance

• Lot number matches container label and purchase order
• Material is within shelf life or retest date
• Product name, grade, and CAS/INCI match purchase specification
• All required test parameters are present per your specification
• Results are within your specification limits (not just the supplier’s)
• Test methods are cited by name and number for each parameter
• Units are stated and consistent with your specification
• Reporting limits for trace analytes are below your specification limits
• Microbial results cover all required organisms per applicable standard
• Laboratory is ISO 17025 accredited for the tests listed
• CoA is signed and dated by a qualified analyst
• Any narrow-margin passes or unusual results are flagged for trend review

A CoA review is not a rubber stamp. It is an active quality decision. The time spent reading it critically is far less than the time spent managing a recall or regulatory inquiry after a problem reaches the market.