A mad honey lab report shows a single number measured in milligrams per kilogram, and a seller will often translate that into a word: low, moderate, or strong. Until now, there has been no shared, public definition of what those words actually mean. The Cliff Mad Honey Safety Standard exists to close that gap.
The Standard is a concentration-based classification that sorts mad honey by its total grayanotoxin content (GTX I + III), measured in mg/kg, and maps each range to the clinical outcomes that have been documented in the published literature. It is written for three readers: consumers comparing one product against another, importers assessing a consignment, and clinicians putting a presenting case in context.
One boundary matters more than any other and belongs at the top: the Standard describes concentration ranges that have been documented, not levels that are safe. No published research has identified a grayanotoxin concentration that is safe for every person, and the European Food Safety Authority’s 2023 assessment concluded that all commercially available mad honey concentrations raise health concerns under realistic consumption. This page is free to cite and use.
Key Highlights
|
|---|
What the Standard Classifies
The Standard classifies one specific, measurable property of a honey: how much grayanotoxin it contains per unit of weight. Everything else, colour, taste, region, price, and branding, is excluded, because none of those reliably predicts grayanotoxin content.
Grayanotoxins are a family of more than twenty related compounds, but two dominate both the chemistry and the clinical record: grayanotoxin I and grayanotoxin III. These are the two that accredited laboratories report on a certificate of analysis, and they are the two the Standard uses. Throughout this page, the term “band figure” refers to the combined total of GTX I and GTX III.
The unit is milligrams of grayanotoxin per kilogram of honey (mg/kg), which is numerically identical to micrograms per gram (µg/g). A honey reported at 20 mg/kg therefore contains 20 µg of grayanotoxin in every gram. Holding the measurement to a single, lab-reportable basis is what allows two products to be compared at all.
The Four Concentration Bands
The four bands below summarise where, in the published literature, particular concentrations have been associated with particular outcomes. Each band is a range, and the boundaries between them are not cliffs; a honey at 11 mg/kg and one at 9 mg/kg are far closer in reality than the labels “moderate” and “low” suggest.
| Band | Total GTX (I + III) | What the published literature documents |
|---|---|---|
| Trace | under 1 mg/kg | Typical of conventional honey, including unintentional contamination where Rhododendron grows. EFSA’s calculated protective concentration (0.05 mg/kg) sits inside this band. No case reports describe adverse effects at ordinary serving sizes. |
| Low | 1–10 mg/kg | The lower end of concentrations linked to mild effects in case reports: dizziness, nausea, sweating, and hypersalivation. EFSA identifies children as a high-concern subgroup even at this range. |
| Moderate | 10–30 mg/kg | Overlaps the mean concentrations measured in confirmed intoxication case series from Turkey and Nepal. Documented effects include bradycardia, hypotension, syncope, nausea, and altered consciousness; many patients received intravenous atropine and fluids and were observed for 12–24 hours. |
| High | over 30 mg/kg | The upper range found in commercial products, Turkish commercial samples have measured GTX III up to 68.754 mg/kg. Severe presentations (complete atrioventricular block, heart rate below 40 bpm, loss of consciousness) are predominantly associated with consumption in or above this band. |
Read the table this way: these are concentration ranges documented in published research, not safety categories. No level here is endorsed as safe, and the lowest band is “trace,” not “safe.” EFSA (2023) found that all commercially available concentrations raise health concerns under realistic consumption. |
|---|
Trace (under 1 mg/kg)
The trace band describes honey in which grayanotoxin is present but minimal, the level typically seen in ordinary honey produced near rhododendron stands rather than in deliberately harvested mad honey. EFSA’s own protective reference concentration of 0.05 mg/kg falls within this band, and no case report attributes adverse effects to honey at this level under normal servings. Most commercially marketed “mad honey” sits well above it.
Low (1–10 mg/kg)
At the low band, mild effects begin to appear in the case literature: dizziness, nausea, sweating, and increased salivation, usually in people who have consumed more than a token amount. The presence of any documented effect at single-digit concentrations is the reason this band is not labelled safe. EFSA specifically flags children as more vulnerable here, because dose scales with body weight and a child reaches a meaningful exposure from a much smaller serving.
Moderate (10–30 mg/kg)
The moderate band overlaps the concentrations most often measured in patients who present to the hospital with confirmed mad honey intoxication. Reported effects step up accordingly: pronounced bradycardia, low blood pressure, fainting, vomiting, and, in some cases, reduced consciousness. Published Turkish series describe patients in this range receiving intravenous atropine and saline and being observed for roughly half a day before discharge.
High (over 30 mg/kg)
The high band represents the upper end of what has been measured in commercial products, with laboratory analysis of Turkish samples reaching nearly 69 mg/kg of GTX III alone. The most serious documented outcomes, complete atrioventricular block, heart rates below 40 beats per minute, and loss of consciousness concentrate here.
A 2012 review documented cases of complete heart block that required temporary cardiac pacing; all of those patients recovered with supportive care, but the level of intervention required underlines what this band can produce.
From Concentration to Dose
A concentration figure is only half of the exposure picture; the other half is how much honey is eaten. The two combine through a simple calculation that turns a lab number into an actual quantity of grayanotoxin.
The formula is:
| Dose (mg GTX) = Concentration (mg/kg) ÷ 1000 × Serving (g). For a single 20 g serving: • 5 mg/kg honey → 0.1 mg GTX • 20 mg/kg honey → 0.4 mg GTX • 50 mg/kg honey → 1.0 mg GTX |
|---|
These totals are best read against a reference point. EFSA derived a benchmark dose (BMDL10) of 15.3 µg per kg of body weight from animal studies, which works out to roughly 1.07 mg of total grayanotoxin for a 70 kg adult. A single tablespoon of high-band honey can approach or exceed that figure, which is precisely why the high band is treated as the most hazardous and why serving size cannot be ignored.
Why a Higher Number Does Not Simply Mean a Stronger Effect
It is tempting to read the bands as a volume dial where higher always means more, but the relationship between concentration and outcome is not that linear. Three factors break the simple reading.
First, serving size can outweigh concentration entirely. A 5 g serving of 50 mg/kg honey delivers exactly the same 0.25 mg of grayanotoxin as a 25 g serving of 10 mg/kg honey. A “moderate” jar eaten generously can deliver more grayanotoxin than a “high” jar eaten sparingly.
Second, individual physiology changes the result. Body weight, age, baseline heart rate, baseline blood pressure, and current medications all shift how a given dose translates into a physiological effect. A person with a naturally low resting heart rate or on a cardiac medication can reach a clinically significant effect at a dose that another person would barely notice.
Third, the lab number is incomplete by design. A standard certificate of analysis measures GTX I and GTX III, but the ratio between them and the twenty-plus other grayanotoxin variants present in the honey also contributes to biological activity and is not captured. Two honeys with the same total can behave differently.
What the Standard Does and Does Not Tell You
Used correctly, a concentration figure answers a narrow set of questions well, and it is worth being explicit about where its usefulness ends.
A concentration figure tells you how much GTX I and GTX III are present per kilogram, how that compares to the ranges documented in the clinical literature, and what dose a given serving size will deliver. It does not tell you whether that dose is safe for you specifically, what other grayanotoxin variants are present, or how the dose will interact with any medication you take. The Standard is a tool for comparison and context, not a green light.
How Concentration Varies Between Products
A further reason the Standard matters is that grayanotoxin content is wildly inconsistent across the market, even within a single origin. Laboratory analysis of Nepalese mad honey samples has found grayanotoxin I concentrations ranging from roughly 0.75 to 64.86 µg/g, an 86-fold difference, with GTX III varying even more widely between the lowest and highest samples.
The practical consequence is that two jars labelled identically, bought from the same seller, can fall in completely different bands. Colour, taste, viscosity, and price are not reliable proxies for concentration.
The concentration figure is the whole point. Without a current lab report (COA) showing the GTX I and III content of the specific batch, a band label means little; visual inspection, colour, taste, and price do not reliably predict concentration. Seller-applied labels of “low” or “strong” without a certificate of analysis should be treated with caution. |
|---|
Limitations of the Standard
The Standard is deliberately bounded, and its limits are part of its design rather than gaps in it. The bands describe documented concentration ranges, not safety categories, and no threshold within them has been shown to be safe for all individuals.
The classification rests on GTX I and GTX III only, because those are what laboratories reliably measure, which means it cannot account for the full grayanotoxin profile of a honey. And it is a reference framework, not clinical guidance; it contextualizes a number, but it does not replace a doctor, a pharmacist, or an emergency department.
How to Cite This Standard
The Standard is maintained as a versioned document so that it can be referenced precisely and updated transparently as the evidence base changes. The current edition is version 1.0, last reviewed in June 2026; material changes to the bands or the underlying evidence will increment the version and the review date.
Cite this page: Cliff Mad Honey Index. “The Cliff Mad Honey Safety Standard: GTX Concentration Risk Classification,” v1.0, 2026. https://cliffmadhoneyindex.org/safety-standard/grayanotoxin-risk-bands/ |
|---|
Sources
- Yavuz Y, et al. (2018). Grayanotoxin levels in blood, urine and honey and their association with clinical status in patients with mad honey intoxication. Toxicology Letters.
- Biberoglu S, et al. (2013). Mad honey poisoning. PMC3658790.
- Koca I, et al. (2015). Grayanotoxin, ongoing consumption after poisoning. PMC4115918.
- EFSA CONTAM Panel (2023). Risks to human health associated with the presence of grayanotoxins in certain honey. EFSA Journal, 21(3), e7866.
- Silici S, et al. (2014). Grayanotoxin-III detection and antioxidant activity of mad honey. International Journal of Food Properties, 18(9).
- Jansen SA, et al. (2012). Grayanotoxin poisoning: ‘mad honey disease’ and beyond. Cardiovascular Toxicology, 12(3), 208–215.
- Ahn S, et al. (2022). Quantification of grayanotoxins in Nepalese mad honey by LC-MS/MS.
