1Three questions, three groups of tests

In this article, a lot means one batch of honey that is handled as a unit, such as one harvest, one tank, one drum shipment or one consignment. Each of the three questions has its own group of tests:

• Is it honey? The composition tests measure moisture, sugars, HMF, diastase and the other parameters that define honey in law. Pollen analysis supports a floral name on the label.
• Is anything banned in it? The veterinary residue tests look for antibiotics and for the drugs that have no permitted level in honey, such as chloramphenicol, the nitrofurans and the nitroimidazoles.
• What did the bees bring in? These tests look for pesticides from treated crops and from treatments used in the hive, for lead and, in honey from some countries, for radioactive caesium.

Two terms appear throughout the article. A maximum residue limit (MRL) is the highest level of a residue that the law permits in a food. The market is the country where the honey is sold, and its rules set the limits that apply to each result.

A composition result says nothing about residues, and a residue result says nothing about added syrup.

2The tests, one by one

Every test below is listed, with its price, on our honey testing page. The accreditation status of each test is stated in one of three forms:

• Accredited, honey named: the accreditation scope of the method names honey.
• Accredited, food: the method is accredited for foodstuffs or for products of animal origin, a category that includes honey without naming it.
• Not accredited: the method is a validated in-house method or a screening method. Its result is reported without the accreditation mark, so it is not suitable where an accredited result is required.

Is it honey? Composition and quality

• HON-01 composition panel. It measures moisture, reducing sugars, sucrose, water-insoluble matter, electrical conductivity, free acidity, diastase activity and HMF, which are the parameters that define honey in EU law and in the Codex standard. The result shows whether the lot meets the composition criteria of your market. Accredited, honey named, for six of the eight parameters; conductivity and water-insoluble matter are not accredited.
• HON-09 sugar profile. It measures glucose, fructose and sucrose separately by HPLC, so the result gives the proportion of each sugar in the honey. Accredited, honey named.
• HON-10 invertase. It measures invertase, another honey enzyme. The result is read together with diastase and HMF as an indication of freshness and heating. Not accredited.
• HON-02 pollen analysis. It counts the pollen grains in the honey under the microscope. The result shows the botanical origin of the honey, which supports a floral name on the label, and it can also show whether the pollen is consistent with a declared region. Not accredited.
• HON-08 microbiological quality. It measures the total aerobic count, yeasts and moulds, and coliforms, to show whether the honey is clean enough to sell. Retailers ask for these results in their technical questionnaires. Accredited, food.
• HON-12 Fiehe reaction. It is a qualitative colour test for HMF, used as a first check for added invert sugar. A positive result means that a full authenticity test is needed, but a negative result is not proof that the honey is authentic. Not accredited; a screening test.

Is anything banned in it? Veterinary medicine residues

• HON-04 antibiotic residues. It looks in one run, by LC-MS/MS, for the classes of antibiotic used in beekeeping: penicillins, cephalosporins, tetracyclines, quinolones, macrolides, sulfonamides and others. The quotation lists every compound covered, and the result shows whether any of them is present and at what level. Accredited, honey named.
• HON-13 chloramphenicol, confirmatory. It measures chloramphenicol by LC-MS/MS, the reference method. Importers and border authorities ask about this residue specifically, and the result is compared with the EU reference point of 0.15 µg/kg or with the rule of your market. Accredited, honey named.
• HON-05 chloramphenicol, screening. It screens for chloramphenicol by ELISA and gives a quicker answer than HON-13. A positive screen is confirmed by HON-13 before it is reported as a finding. Accredited, honey named.
• HON-14 aminoglycosides. It measures streptomycin and the other aminoglycosides by LC-MS. They need this separate test because a multi-residue antibiotic screen does not cover them. Accredited, products of animal origin.
• HON-15 nitrofurans. It measures the nitrofuran metabolites by LC-MS/MS. Nitrofurans may not be given to food-producing animals, and the EU reference point is 0.5 µg/kg for each metabolite. Accredited, products of animal origin.
• HON-16 nitroimidazoles. It measures metronidazole, dimetridazole and their metabolites by LC-MS/MS. These drugs have no permitted level in honey, and the EU has set no reference point for them. Accredited, honey named.
• HON-06 varroa treatment residues. It measures amitraz, coumaphos and up to 250 other pesticides by LC-MS/MS, to show whether the beekeeper's own varroa treatments have reached the honey. The EU limits are 0.2 mg/kg for amitraz and 0.1 mg/kg for coumaphos. Not accredited for honey.

What did the bees bring in? Pesticides and contaminants

• HON-07 pesticide residues. It screens for many pesticides at once, including the neonicotinoids, by QuEChERS extraction and LC-MS/MS, and the quotation lists every compound covered. The results are compared with the honey MRLs of your market; in the EU, a pesticide that has no specific MRL for honey is held to 0.01 mg/kg. Accredited, food.
• HON-03 heavy metals. It measures lead, cadmium, mercury and arsenic by ICP-MS. The lead result is compared with the limit of your market, which is 0.10 mg/kg in the EU, and the other three metals are reported alongside it. Accredited, food.
• HON-11 radioactivity. It measures caesium-134 and caesium-137, iodine-131, ruthenium-103 and 106, and americium-241 by gamma spectrometry. For honey from the countries listed in the EU Chernobyl rules, the result shows whether the lot is within 600 Bq/kg of caesium-137. It is the fastest result on the list. Accredited, food.

None of these tests proves that a honey is free of added syrup. Section 4 explains which methods can detect syrup, and why a composition result cannot.

3Is it honey? The composition rules

In the EU, honey is defined by Directive 2001/110/EC. Annex II of the Directive lists the composition criteria that honey must meet when it is sold:

• Fructose plus glucose: at least 60 g per 100 g, or at least 45 g for honeydew honey and for blends of honeydew honey with blossom honey.
• Sucrose: at most 5 g per 100 g. The limit is 10 g for false acacia, alfalfa, citrus and a few other named honeys, and 15 g for lavender and borage.
• Moisture: at most 20 %, or 23 % for heather honey and for baker's honey in general, and 25 % for baker's honey from heather.
• Water-insoluble matter: at most 0.1 g per 100 g, or 0.5 g for pressed honey.
• Electrical conductivity: at most 0.8 mS/cm, but at least 0.8 mS/cm for honeydew and chestnut honey. No value is set for strawberry tree, bell heather, eucalyptus, lime, ling heather, manuka and tea tree honey.
• Free acid: at most 50 milli-equivalents per kg, or 80 for baker's honey.
• Diastase activity: at least 8 on the Schade scale. Honeys with a low natural enzyme content, such as citrus honey, need only 3, provided that their HMF is not above 15 mg/kg.
• HMF (hydroxymethylfurfural): at most 40 mg/kg, or 80 mg/kg for honey of declared origin from regions with a tropical climate.

Diastase and HMF are measured after processing and blending. The Annex also states that nothing may be added to honey, and that pollen and other constituents may not be removed unless this is unavoidable when foreign matter is removed.

England, Wales and Scotland apply the same values to the same named honeys, and so do Norway and Switzerland.

Our tests: HON-01 composition panel, HON-09 sugar profile.

Outside Europe

Codex. Codex Alimentarius is the international food standards programme of FAO and WHO. Its honey standard, CXS 12-1981, sets the same limits as the EU for sugars, sucrose, moisture and water-insoluble matter. Diastase, HMF, free acidity and conductivity appear only in an Annex, which the standard describes as intended for voluntary use by commercial partners rather than by governments.

Canada. Canada grades honey, and grading is mandatory for honey that is imported, exported or sent from one province to another. The moisture limit depends on the grade: 17.8 % for Canada No. 1 (18.6 % if the honey is labelled pasteurized), 18.6 % for No. 2 and 20 % for No. 3. The grade standard also sets limits for reducing sugar, sucrose, water-insoluble solids, free acid and ash, which Canada uses instead of conductivity. Diastase must be at least 8 on the Gothe scale when HMF is not above 40 mg/kg, or at least 3 when HMF is not above 15 mg/kg; there is no separate HMF limit.

Australia and New Zealand. Australia sets only two criteria, in Standard 2.8.2 of the Food Standards Code: at least 60 % reducing sugars and at most 21 % moisture. It has no criterion for sucrose, HMF or diastase. The Code is shared with New Zealand. Standard 2.8.2 is not marked as applying to Australia only, but we found no New Zealand text that confirms it applies there.

United States. The United States has no federal standard of identity for honey. The FDA states that honey with any other sweetener added must be labelled as a blend, not as honey, and must declare the added sweeteners.

Sources 1, 3, 13–15, 17, 19, 23, 27.

What HMF and diastase can show

HMF and diastase are used to judge whether honey has been heated or stored for too long, but neither value proves this on its own.

HMF forms in honey during storage. In one study, fourteen commercial honeys were kept at room temperature for up to eighteen months: HMF rose by about 3 mg/kg a month in citrus and eucalyptus honey, and by about 0.26 mg/kg a month in chestnut honey.

Diastase, by contrast, falls during storage. In the same study it did not change during the first eight months, and then fell at rates that differed from one type of honey to another. Some honeys are also naturally low in diastase, and the EU accepts a value of 3 instead of 8 for them.

No EU, UK or Codex text sets a limit for invertase. The amended EU Directive, however, asks the Commission to set criteria on heating and enzymes, "taking into account the invertase index", by 14 June 2029.

Our tests: HON-01 composition panel, HON-10 invertase.

Sources 1, 33.

HMF and diastase should be read together, with the harvest date and the storage history of the lot. A high HMF or a low diastase on its own does not show that the honey was heated.

What pollen analysis can show

Pollen analysis identifies the pollen grains in a honey under the microscope. The harmonised method counts at least 500 grains, and the result is read as follows:

• Floral origin: a honey is taken to come mainly from one plant when that plant's pollen makes up more than 45 % of the count. Because some plants give far more pollen than others, the method also takes account of taste, smell and physico-chemical results.
• Geographical origin: the method checks whether the whole pollen spectrum is consistent with the flora of a region.
• Limits: precision is very poor for pollen types present at low frequencies, and the result depends on the experience of the analyst.

Our test: HON-02 pollen analysis.

Source 34.

Pollen analysis supports a floral name on the label and can test a declared region against its flora, but the method gives no way to measure the share of each country in a blend.

4Can a test show that honey was mixed with syrup?

The composition tests in section 3 do not reliably detect added syrup. In the EU's most recent coordinated action, Member States reported that most of their official laboratory techniques were deficient in detecting and proving adulteration with extraneous sugars when used alone.

What the authorities found

These are the most recent official programmes that tested honey for added sugar:

• EU, 2021 to 2022. Sixteen Member States, Norway and Switzerland sampled 320 consignments of imported honey at the EU border between November 2021 and February 2022. The Commission's Joint Research Centre found 147 of them (46 %) suspicious, meaning that a marker of adulteration was detected or that the carbon isotope ratio was outside the values of authentic honey. By declared origin, 66 of the 89 consignments from China, 14 of the 15 from Türkiye and all 10 analysed from the United Kingdom were suspicious; the authors note that the UK consignments may have been honey produced elsewhere and processed in the UK. Of the 96 EU importers that received the sampled consignments, 63 had received at least one suspicious consignment.
• EU, 2015 to 2017. In an earlier EU control plan, the Joint Research Centre tested 893 samples and found added sugar in 14 % of them.
• Canada, 2024 to 2025. The Canadian Food Inspection Agency tested 81 honey samples taken at businesses that it had selected as higher risk. Fifteen of them (19 %) were unsatisfactory for foreign sugars, and all fifteen were imported: 15 of the 68 imported samples, against none of the 13 Canadian ones. The agency prevented 133,420 kg of adulterated honey from being sold, and it notes that targeted results do not represent the market as a whole.
• United States. The FDA found undeclared sweeteners in 14 of 144 imported samples in 2021–22 (10 %), in 3 of 107 imported samples in 2022–23 (3 %) and in 4 of 102 domestic and imported samples in 2025 (4 %). It notes that the three assignments were not designed for statistical comparison, and that its current method detects only a subset of undeclared sweeteners.
• Australia. Imported honey is classed as a surveillance food, and 5 % of consignments are referred for testing for C4 sugar (at most 10 %), moisture and reducing sugars. We found no published results.

A review of the EU action's methods, published by Defra in 2025, stresses that "suspicious" does not mean proven adulteration.

Sources 22, 27–32.

Every programme that published results found honey that did not pass its authenticity tests. The shares cannot be compared, however, because each programme chose its samples differently and used different methods.

Which methods detect syrup

The EU action used three kinds of method:

• Carbon isotope ratio (EA-IRMS and LC-IRMS). Maize and sugar cane belong to a group of plants called C4 plants, and a syrup made from them shifts the carbon isotope ratio of the honey strongly. Syrups from sugar beet, rice or wheat, which are C3 plants, shift it less. The method compares the ratio in the protein of the honey with the ratio in each of its sugars, which allows it to detect C4 syrups and also C3 syrups. Measuring the sugars alone is not enough to confirm a C4 syrup.
• Marker compounds, by high-resolution mass spectrometry (LC-HRMS). The action looked for four kinds of marker: mannose, which points to starch-based syrups; difructose anhydride III, which points to high-fructose corn syrup and inverted syrup; 2-acetylfuran-3-glucopyranoside, which points to rice syrup; and long-chain sugars from starch-based syrups. Mannose also occurs naturally in some honeys, such as linden, eucalyptus and chestnut honey.
• NMR profiling. The spectrum of the honey is compared with a commercial database of authentic honeys. The action used it as supporting evidence, and its sensitivity is lower than that of LC-HRMS.

The techniques did not agree on every sample, and the authors state that the methods have not yet been validated between laboratories. The amended EU Directive requires the Commission to lay down, by 14 June 2028, the methods of analysis for detecting adulterated honey.

We do not currently offer isotope-ratio, NMR-profiling or LC-HRMS tests for honey authenticity.

Sources 1, 27, 30.

The Fiehe reaction

The Fiehe reaction is an older colour test for HMF. HMF forms from fructose when sucrose is split with acid to make invert sugar, so the test is used to look for added invert syrup.

HMF, however, does not prove adulteration. A 2019 review states that HMF cannot be used alone as an indicator of adulteration, and a 2009 study found that honey mixed with corn syrup or invert syrup of known origin remained within the EU limits for HMF.

Our test: HON-12 Fiehe reaction.

Sources 35–37.

A positive Fiehe result is a reason to order an authenticity test, while a negative result proves nothing.

Blends: the new EU label

Since 14 June 2026, EU rules require the label of a blend of honeys from more than one country to name every country of origin. The countries must appear in the principal field of vision, in descending order of their share by weight, each with its percentage. A tolerance of 5 % is allowed on each share.

• On packs of less than 30 g, the country names may be replaced by two-letter country codes.
• When a blend comes from more than four countries and the four largest shares make up more than half of it, a Member State may allow percentages for the four largest countries only.
• Honey placed on the market or labelled under the old rules before 14 June 2026 may be sold until stocks run out.

The same amendment removed "filtered honey" as a category. Honey filtered in a way that removes a significant share of its pollen now counts as baker's honey, which must be labelled "intended for cooking only".

The other markets covered here follow different rules:

• Northern Ireland applies the EU rule from 14 June 2026.
• England, Wales and Scotland still allow a blend to be labelled "blend of honeys from more than one country", without the countries or their percentages, and they keep "filtered honey" as a category.
• Norway will apply the amendment only once it has been taken into the EEA Agreement, and the Norwegian Food Safety Authority gives no date for this. Until then, a blend may still be labelled as honey from the EU, from outside the EU, or from both; for this purpose, Norway counts as outside the EU.
• Switzerland allows a blend to name a wider area, such as "EU" or "South America", instead of the countries.
• Canada requires a blend to be labelled "A Blend of Canadian and (country) Honey", with the countries in descending order of their share but without percentages.
• Codex requires the country of production only when a honey is named after its floral source or region, and it has no rule for blends.
• The United States applies its general rule that imported goods must show their country of origin, and we found no rule specific to honey. New Zealand's country-of-origin rules do not cover honey, and we could not read the Australian origin standard.

Sources 1–4, 13–15, 17, 21, 26.

The percentages on an EU blend label are calculated from the operator's traceability documents, not from a test. The Commission is to set methods and criteria for identifying the place of harvest, and for traceability, by 14 June 2029.

5Is anything banned in it?

Veterinary medicines are used in hives against bee diseases and against the varroa mite, and each market in this article regulates their residues in its own way.

The EU, the UK, Norway and Switzerland

In the EU, a veterinary medicine may be given to a food-producing animal only if its residues have been classified under Regulation (EC) No 470/2009. Bees count as food-producing animals, and Regulation (EU) No 37/2010 sets the limits that apply to their honey:

• Only two substances have a numerical MRL in honey, and both are varroa treatments: amitraz at 200 µg/kg and coumaphos at 100 µg/kg.
• Oxalic acid, flumethrin and tau-fluvalinate are listed for bees with "no MRL required", and formic acid and thymol are listed in the same way for all food-producing species.
• No antibiotic has an MRL in honey.

Table 2 of Regulation 37/2010 lists the substances for which no MRL can be established. These substances may not be given to food-producing animals, and they include chloramphenicol, the nitrofurans (among them furazolidone), dimetridazole, metronidazole and ronidazole.

For some of these prohibited substances, the EU has set a reference point for action (RPA): a food with a residue at or above the RPA is considered not to comply with EU law and may not enter the food chain. Regulation (EU) 2019/1871 sets two RPAs that apply to honey:

• chloramphenicol: 0.15 µg/kg;
• nitrofuran metabolites (AOZ, AMOZ, AHD, SEM and DNSH): 0.5 µg/kg each.

The RPAs apply to food of animal origin "irrespective of the food matrix tested", and Regulation 470/2009 names honey as a food of animal origin.

No RPA has been set for the nitroimidazoles. Regulation 470/2009 treats food with residues of a prohibited substance as non-compliant unless an RPA exists and the residue is below it, so this exception is not available for the nitroimidazoles.

Great Britain applies the same RPA values, Norway applies the EU regulations through the EEA Agreement, and Switzerland refers to them directly.

Our tests: HON-04 antibiotic residues, HON-13 chloramphenicol, confirmatory, HON-05 chloramphenicol, screening, HON-15 nitrofurans, HON-16 nitroimidazoles.

Sources 5–7, 13, 14.

Canada, the United States, Australia and New Zealand

Outside Europe, some antibiotics do have a limit in honey, as the figure shows.

• Canada prohibits the sale of any food from an animal that contains a residue of chloramphenicol, a 5-nitrofuran or a 5-nitroimidazole. It also publishes working residue levels for other antibiotics in honey, such as 0.0375 mg/kg for streptomycin, but these are guidance, not regulation.
• The United States prohibits the extralabel use in animals of chloramphenicol, of the nitrofurans furazolidone and nitrofurazone, and of the nitroimidazoles. Under FDA Import Alert 36-04, honey with residues of chloramphenicol, nitrofurans or fluoroquinolones is detained.
• Australia permits a residue of a veterinary or agricultural chemical in a food only if Schedule 20 of the Food Standards Code lists that chemical for that food. We found no honey entry for chloramphenicol, the nitrofurans or the nitroimidazoles.
• New Zealand's MRL notice sets chloramphenicol at 0.00015 mg/kg (0.15 µg/kg) in any food.

Sources 18–20, 24.

Honey that complies in one market does not automatically comply in another.

Screening and confirmation

Residue tests are of two kinds. A screening method detects a substance, or a group of substances, at the level of interest, while a confirmatory method identifies the substance unequivocally and, where necessary, measures it.

Under the EU rules on residue testing, a suspect screening result must be confirmed by a confirmatory method, and the result is non-compliant when it is at or above the decision limit of that method, known as CCα.

Aminoglycosides, such as streptomycin, need a method of their own. They are very polar and give a weak response in mass spectrometry, which makes them difficult to include in a multi-residue method.

Our test: HON-14 aminoglycosides.

Sources 8, 38.

A positive screen therefore needs confirmation before it counts as a finding, and an antibiotic screen that leaves out the aminoglycosides says nothing about streptomycin.

6What did the bees bring in?

Three things can reach the honey from outside the hive: pesticides from the crops the bees visit, lead from dust around the apiary and, in honey from some countries, radioactive caesium left by the Chernobyl accident. The varroa treatments used inside the hive are included here as well, because they are also regulated as pesticides.

Pesticides and varroa treatments

EU pesticide legislation gives honey its own product code, 1040000. Where no specific MRL is set for a pesticide in honey, a default limit of 0.01 mg/kg applies. Amitraz and coumaphos are regulated twice, as veterinary medicines and as pesticides, with the same values in both cases: 0.2 and 0.1 mg/kg.

The table compares the honey values of five markets, in mg/kg.

Clothianidin and thiamethoxam have the same values as imidacloprid in the EU and Great Britain, and no honey entry in Australia, New Zealand or the United States. We did not read the honey values of Norway, Switzerland or Canada, whose default limits are 0.01, 0.01 and 0.1 mg/kg.

Our tests: HON-06 varroa treatment residues, HON-07 pesticide residues.

Sources 6, 9, 10, 13, 14, 18–20, 25.

The same honey can be within the EU limit for acetamiprid and above the British one, because the EU sets 1 mg/kg and Great Britain 0.05 mg/kg. For thiacloprid the difference runs the other way: 0.05 mg/kg in the EU and 0.2 mg/kg in Great Britain.

Lead

The EU limit for lead in honey is 0.10 mg/kg, and Great Britain, Norway, Switzerland and Codex set the same figure. Australia and New Zealand set no lead limit for honey; their only contaminant limit for honey is 0.7 mg/kg for tutin, a natural toxin. We found no lead limit for honey in Canada or the United States.

Part 4 of our Heavy Metals series explains where the lead in honey comes from and when to sample for it.

Our test: HON-03 heavy metals.

Sources 11, 13, 14, 16, 18, 19.

Radioactivity

Implementing Regulation (EU) 2020/1158 sets the conditions for importing food from the countries affected by the Chernobyl accident: Albania, Belarus, Bosnia and Herzegovina, Kosovo, Moldova, Montenegro, North Macedonia, Russia, Serbia, Switzerland, Türkiye, Ukraine and Great Britain.

• Food from these countries may not exceed 600 Bq/kg of caesium-137, or 370 Bq/kg for milk products and for food for infants and young children.
• An official certificate is required only for wild mushrooms, certain wild berries and products made from them. Honey needs no certificate, but the 600 Bq/kg limit still applies to it.
• Great Britain applies the same limits to food from the same countries, with the exception of the United Kingdom itself.
• Norway repealed its national limits for caesium in food on 1 January 2025. Switzerland sets radionuclide limits for use after a nuclear accident, and we found no Swiss limit for normal conditions.

Our test: HON-11 radioactivity.

Sources 12–14.

For honey from one of the listed countries, a caesium result shows whether the lot is within the EU limit of 600 Bq/kg.

7Which tests for whom

The table matches each type of business with its question and with the bundle on our honey testing page that answers it. Each bundle is run on one set of samples, and its results come back in a single report.

Each result is compared with the limit of the market you name, and where that market sets no limit for a parameter, the report says so. The report is not a declaration that the honey complies with the law.

8The sample

One jar is sent for each lot. It is filled from the lot, closed with a tight lid and marked only with a sample code, with no brand, no origin and no paperwork inside the parcel. The quotation states how much honey to send.

We send you the code list, a prepaid UPS label and the customs documents for laboratory samples, and you drop the parcel at a UPS point in any country from which UPS ships.

The result applies to the sample that was tested, and how well it represents the lot depends on how the jar was filled. Please note the harvest date and the storage history of the lot, because HMF and diastase are read together with both.

The report names the market and the limit that applies to each result, together with the method, the reporting limit and, wherever the method provides one, the measurement uncertainty.

9One table to keep

Name the market before the lot is tested: the same result can meet the limit of one market and not that of another.

Part 3 of the Heavy Metals series explains how to read the result itself, including the reporting limit and the measurement uncertainty.