People keep asking how to evaluate a CoA. Let me break down what to look for in a mad honey CoA specifically:
Header section:
- Lab name and accreditation (ISO 17025 is the relevant standard)
- Report date
- Sample ID (this is what you match to your order's batch number)
- Product description (should reference honey and grayanotoxin/diterpenoid testing)
Results section:
- Grayanotoxin value in mg/kg — what you're actually there for. Look for a specific number, not just "detected" or "compliant"
- Which GTX variants were tested (GTX-I and GTX-III are most relevant)
- Detection limit listed — if the value is near the detection limit, precision is lower
- Method: HPLC is standard; HPLC-MS is better
Red flags:
- No specific grayanotoxin number, just "compliant" or "tested"
- Lab name you can't find with a basic search
- Date that predates the claimed harvest season
- No sample ID, or sample ID that doesn't match your order
Good guide. One addition: the difference between HPLC and HPLC-MS (mass spectrometry) matters. Plain HPLC can measure total content reasonably well but HPLC-MS gives you better specificity for individual GTX variants and is significantly harder to fake convincingly. If you see an HPLC-MS CoA from an ISO 17025 accredited lab with a named analyst, that's as good as it gets in terms of documentation quality.
What's a typical grayanotoxin number for genuine high-quality honey? Trying to understand what the figures mean in practice.
Potency varies enormously by harvest and source but some rough orientation:
- Low potency (autumn harvest, lower altitude): 0.5–2 mg/kg
- Moderate (typical commercial product): 2–8 mg/kg
- High potency (spring cliff harvest, Apis dorsata laboriosa, Mustang/Lamjung): 8–20+ mg/kg
If a seller shows you a CoA but won't give you the actual grayanotoxin figure, or the document says "tested and compliant" without a number, that's a red flag. A real CoA has a real number. "Tested" is not a result.