Raw vs Processed Honey: What Heat and Filtering Really Change, and What They Do Not

Few food labels carry as much confidence and as little definition as "raw honey". Shelves are full of jars promising raw, natural, pure, unfiltered or unheated honey, usually at a higher price, and shoppers reasonably want to know whether the difference is real. The honest answer has three parts. Yes, heating and filtering change honey in ways a laboratory can measure. No, those changes do not turn honey into a health food, and the sugar in raw honey counts exactly like the sugar in any other honey. And the most important quality question is often not raw versus processed at all, but whether the jar contains only honey.

This guide goes through what "raw" does and does not mean, what the international honey standard actually requires, what the best available comparison of raw and imported honey found (including where it did not support the popular story), what heat itself does in temperature experiments, how honey relates to blood sugar, and how to buy honey when the label alone cannot tell you what is inside. We have corrected several figures that appear in older versions of this page, and we say where.

What "raw honey" means, and what nobody regulates

Ask five sellers what raw honey is and you may hear five answers. The National Honey Board in the United States defines it as honey as it exists in the beehive or as obtained by extraction, settling or straining without adding heat. In practice, the food trade and state agriculture departments generally understand raw honey to mean honey that has not been pasteurised and has not been micro-filtered in a way that removes most of the pollen. Neither description is a legal standard. There is no federal standard of identity that defines raw honey, and, as industry commentators point out, no authority checks the words "raw", "natural" or "unfiltered" before a jar reaches the shelf. In other words, "raw" is a promise made by the seller, not a status verified by the state.

The international honey standard, the Codex Alimentarius Standard for Honey (CXS 12-1981), does not use the word raw either. What it does say is more useful. Honey is defined as the natural sweet substance produced by Apis mellifera bees from the nectar of plants or from secretions of living parts of plants or excretions of plant-sucking insects, which the bees collect, transform, deposit, dehydrate, store and leave in the comb to ripen and mature. Honey sold as such must not have any food ingredient added to it, including food additives, and honey "shall not be heated or processed to such an extent that its essential composition is changed and/or its quality is impaired". That last clause is the closest thing the international standard has to a rule about raw versus processed: heating is allowed, but not to the point of damaging the product, and the standard supplies measurable markers to police it.

The Codex standard also has a labelling rule that matters for the raw debate. Where pollen has been significantly removed, the product must be designated "filtered honey". Pollen is one of the tools scientists use to verify a honey's botanical and geographic origin, so filtration that removes it is not just a cosmetic choice. A jar labelled simply honey is not expected to have had its pollen significantly removed.

How honey gets from the hive to the jar

Honey is made in the hive and ripened by the bees, which fan the nectar until most of the water has evaporated and cap the cells with wax. A beekeeper removes the frames, cuts or scrapes off the caps and extracts the honey, typically by spinning the frames in a centrifuge so the honey is flung out of the comb. The liquid is then strained through mesh to remove wax and hive debris and is often left to settle so that air bubbles and fine particles rise to the top. Honey handled only in this way, and bottled at hive temperature, is what most people mean by raw.

Larger packers add steps. Honey is often warmed so that it flows easily through pipes and filters, and to dissolve the fine sugar crystals that would otherwise make it set. It may be filtered finely to give a clear, slow-to-crystallise product that looks the same in every jar, and honeys from different batches, seasons or countries may be blended to a standard colour and flavour. Each of these steps has a rationale that is commercial rather than sinister: consumers in many markets prefer clear, liquid honey that stays liquid for months on a shelf. The question is what each step costs in quality, and the laboratory markers in the next section are how that cost is measured.

The markers laboratories use

Moisture. Ripe honey contains less than about a fifth water. Higher moisture allows yeasts to ferment the honey, which is why the standard caps it at 20% (23% for heather honey).

Diastase activity. Diastase is an enzyme naturally present in honey. Its activity, measured in Schade units and often called the diastase number, is a marker of freshness and gentle handling, because the enzyme loses activity when honey is heated or stored for a long time. The Codex minimum is 8 Schade units in general and 3 for honeys with naturally low enzyme content.

Hydroxymethylfurfural (HMF). HMF is a compound that forms as honey sugars break down over time and with heat. Fresh honey has little of it, and the level rises with heating and long storage, which makes it a marker of heat exposure and age. Codex allows up to 40 mg per kilogram, or 80 for honeys from tropical countries, where warmer ambient conditions raise HMF more quickly.

Sugar profile and sucrose. Genuine honey is mostly fructose and glucose, with only a small amount of sucrose. Unusual sugar ratios are one of the classic signals that honey may have been diluted with cheaper syrups or that bees were fed sugar.

None of these markers is a health claim. They are quality and authenticity measures, and a honey can pass every one of them and still be as high in sugar as any other. Table 1 sets out the main Codex requirements and what each is designed to catch.

What the best raw-versus-imported comparison found

The claim that raw honey differs from commercial honey is often repeated and rarely tested, so it is worth looking at one of the few studies that did. In 2021, a team at the University of Rzeszów in Poland published a comparison in the journal Molecules of five raw local Polish honeys (acacia, buckwheat, linden, multifloral and honeydew, from ecological apiaries) and eleven imported honey blends sold on the Polish market, labelled as mixtures of EU and non-EU honeys. The authors measured antioxidant activity by three methods, total phenolic content, soluble protein, diastase and acid phosphatase enzyme activity, and standard quality parameters including water, HMF, acidity and conductivity. They were motivated by a concern that many imported honeys compete with local honey mainly on lower price, which can correspond to lower quality and widespread adulteration.

The headline result was clear. Raw local Polish honeys showed higher antioxidant and enzymatic activity and more soluble protein. Diastase activity ranged from 5.63 to 66.43 in the local honeys and from 2.7 to 20.7 in the imported blends. Soluble protein ranged from about 38 to 475 mg per 100 g in the local honeys against about 11 to 225 mg per 100 g in the imports. Total phenolic content, a common antioxidant marker, ranged from about 141 to 4,973 mg gallic acid equivalents per kg in local honeys and from about 59 to 653 in the imports, and the FRAP antioxidant measure was likewise substantially higher in the local samples. Protein content and diastase number were the two parameters that best separated imported blends from raw local honeys statistically.

Now the caveats, which matter as much as the result. The HMF finding was mixed, not one-directional: HMF ranged from 12.5 to 22.66 mg/kg in the local honeys and from 6.4 to 73.92 mg/kg in the imports, so the imported group contained both the lowest and the highest values, and two imported samples exceeded the 40 mg/kg Codex limit. Older summaries of this study, including an earlier version of this page, described imported honey as simply carrying elevated HMF. That was too strong. The study was small (16 samples), from one region of one country, and it compared raw local monofloral honeys with blended imports, so it cannot tell you how much of the difference came from heating, how much from age and storage, how much from the botanical source, and how much from blending. The authors themselves suggest thermal processing of commercial honey as a likely reason for reduced enzyme and antioxidant activity, but they did not isolate it.

What heat does on its own

Because the Polish study could not separate heating from everything else, it helps to look at experiments that heat the same honey and measure what changes. A study published in the Journal of Food Science and Technology in 2016 took raw rape honey (Brassica napus) and heated it at temperatures from 30 to 80 degrees Celsius. It found that diastase activity fell with heating at higher temperatures, and that HMF rose most in honey heated at 80 degrees Celsius for 15 minutes, by 62 percent on average compared with unheated samples. Yet the authors also concluded that heating for 15 minutes between 50 and 80 degrees Celsius did not significantly degrade the quality of the honey, and only slightly increased HMF and reduced diastase.

That is a more moderate message than the internet's "heat destroys honey". Short warming to a liquefying temperature has a measurable but small effect on these markers. What matters more is the combination of temperature, time and storage: honey that is heated repeatedly, held hot for long periods, or stored warm for months will accumulate HMF and lose diastase. That is exactly the pattern the Codex markers are meant to expose.

A 2022 study in Food Chemistry looked at the opposite problem: how to keep honey liquid for a year without ruining it. The researchers used HMF, diastase activity, acidity and total phenolic compounds as quality markers, tested different treatments in different honey types, and concluded that only storing samples at minus 18 degrees Celsius kept the nutritional benefits of honey and gave results similar to fresh, unheated samples. In plain terms, the treatments that keep honey clear on a shelf have a quality cost, and the honey that keeps its markers best is honey that is allowed to crystallise or is frozen.

Does raw honey do more for your health?

This is the question most buyers actually care about, and the evidence is thinner than the marketing suggests. The laboratory markers above, and the antioxidant tests in the Polish study, describe the chemistry of the honey in a test tube. They do not by themselves show that eating raw honey produces better health outcomes than eating processed honey, and they are not clinical trials.

The nearest we have to clinical evidence is the 2023 meta-analysis in Nutrition Reviews of 18 controlled trials with 1,105 participants, from the University of Toronto. Its authors found significant subgroup differences by floral source and by honey processing, with robinia honey (often sold as acacia honey), clover honey and raw honey showing beneficial effects on fasting glucose and total cholesterol. They concluded that honey, especially robinia, clover and unprocessed raw honey, may improve glycemic control and lipid levels when consumed within a healthy dietary pattern, and that more studies focusing on floral source and processing are needed to increase certainty. Read that sentence twice: the effects were small, the certainty was low, the benefit was measured within a healthy diet and not on top of a poor one, and the authors themselves called for more evidence. It is an interesting signal, not an established advantage.

For laboratory work on raw honey and bacteria, including the difference between the well-known Manuka honeys and other raw honeys and the reasons a laboratory result is not a treatment claim, see our antibacterial honey explainer. The short version is that laboratory activity against bacteria in a dish does not make a jar of honey a medicine.

Honey and blood sugar: what the evidence supports

People with diabetes and prediabetes ask about honey more than almost any other food, and the answer deserves precision rather than a comforting yes or a blanket no. Some of the numbers that circulate online, including some on earlier versions of this page, do not hold up. An earlier version cited an average glycemic index of 58 for honey against roughly 80 for table sugar, and a meta-analysis said to show improved HbA1c at about 10 grams a day. We could not verify those figures in the primary literature, so we have removed them and replaced them with the sources below.

The glycemic index (GI) ranks how quickly a carbohydrate raises blood glucose relative to pure glucose, scored 100. A value of 55 or below is conventionally low, 56 to 69 medium and 70 or above high. Measured honey values sit across that range. A mean of 61 has been cited for honey, which is medium. In a 2023 study in the journal Foods, 26 healthy volunteers were tested with three Mexican honeys, and the glycemic indexes were 74.24 for a multifloral honey (high), 69.20 for a highland honey and 66.36 for an avocado honey (both medium). The honey with the lowest fructose to glucose ratio, the multifloral honey at 1.00, had the highest glycemic index. That is one study in healthy young adults, and an Australian government-funded report from 2005 that tested honeys from different floral sources in volunteers likewise reported wide differences from one honey to another. The safe takeaway is that honey is not reliably "low GI", and that you cannot know the GI of your particular jar without testing it.

Now the clinical trials. A 2021 systematic review of 13 trials in the journal Evidence-Based Complementary and Alternative Medicine, from researchers in Tehran, concluded that oral honey might have no significant effect on metabolic profiles in people without diabetes, and that a high intake might increase glucose and worsen other metabolic parameters in people with type 2 diabetes. Eight of the 13 trials had no control group and eight were of low methodological quality, so the authors urged caution. The Toronto meta-analysis found small favourable changes in fasting glucose, but rated the certainty low and drew mostly on people who did not have diabetes.

Put together, the fair reading is this. Honey is a carbohydrate, and for a person with diabetes it counts toward carbohydrate targets in the same way as sugar, fruit juice or bread. Small, planned amounts are unlikely to be a catastrophe, and honey is not a treatment, a "diabetic" sweetener or a free food. There is no evidence that raw honey is safer than processed honey for blood sugar, because the difference between them is enzymes and pollen, not sugar. How much fits your plan depends on your medication, your targets and your readings, and it is a decision for you and your care team. For a full breakdown of daily amounts and the sugar limits behind them, see our honey dosage guide.

The bigger problem: is it honey at all?

For most shoppers, the raw versus processed debate is a distraction from a more basic question: whether the jar contains only honey. In March 2023 the European Commission published the results of a coordinated action called "From the Hives", in which 16 EU member states plus Switzerland and Norway sampled 320 honey consignments imported from 20 countries between November 2021 and February 2022. The Commission's Joint Research Centre found that 147 samples, 46 percent, were suspicious of being adulterated, meaning they were flagged as non-compliant with EU requirements. Those were suspected rather than confirmed cases, and investigators had to carry out follow-up traceability checks at import, processing, blending and packing sites.

The reported malpractices are instructive. Fraudsters have shifted from maize-based syrups to syrups made from rice, wheat or sugar beet, which better imitate honey's sugar profile, and the price gap is the driver: the JRC cited authentic EU honey at an average of 2.32 euros per kilogram against 0.40 to 0.60 euros for rice syrup. The reported tactics also included blending sugar and honey to a recipe designed to slip past laboratory tests, adding colourings, and masking the true origin of honey by falsifying traceability records and removing pollen. The JRC's own caution deserves quoting: current analytical methods are not yet developed enough at international level, and a simple test is not sufficient to detect adulteration.

Two practical points follow. First, the quick kitchen tests that circulate online, with water glasses, thumbs or flames, should not be trusted to prove purity, since even specialist laboratories find the problem hard. Second, "raw" on a label is not evidence of authenticity. A jar can be labelled raw and be diluted, and a lightly heated honey from a traceable beekeeper can be perfectly genuine. Origin, traceability and honest labelling tell you more than a single word.

Reading a honey label

Table 3 lists the words you are most likely to meet, what each does tell you and, just as usefully, what it does not.

How to read a honey laboratory report

If a seller does supply a laboratory report, four lines matter most, and the Codex limits in Table 1 give you the yardstick. Moisture should be at or below 20 percent, and a value in the high teens suggests a properly ripened honey. HMF should be comfortably under 40 mg per kilogram (80 for tropical origin), and generally speaking values in the low tens or below point to gentle handling and freshness, while values approaching the limit suggest heat, age or both. Diastase activity should be at or above 8 Schade units, or 3 for a floral type that naturally contains little of the enzyme, and a much higher value indicates that the enzyme has been well preserved. Finally, the sugar profile and the sucrose figure should fall inside the ranges the standard sets for that type of honey.

Two cautions apply. First, these routine tests show whether a honey meets quality criteria, not whether it is free of added syrup, since the reported fraud methods include blends designed to pass routine checks. Specialised authenticity analysis exists and is what investigators use, but it is not something a jar's label can summarise. Second, a report is only meaningful if it names the laboratory, the batch and the date, and if you read it as a whole. In the Polish study, nine of the eleven imported blends stayed under the HMF limit, so passing one marker tells you little, and a batch that passes all four looks far better than one that passes only the easy ones.

How to buy honey you can trust

Because a label cannot settle authenticity, the best protection is information you can check. These are the questions worth putting to any seller, whether a supermarket, a shop or a producer like us.

• Where was it produced, and by whom? A named country, region or beekeeper is a good sign. A vague "product of several countries" is not a red flag by itself, but it makes checking impossible.
• Which harvest year and which floral source? Honest sellers can say, and monofloral claims should be backed by pollen or chemical analysis.
• Was it heated or filtered, and how? "Gently warmed to bottle" and "pasteurised" are different answers, and both are acceptable if stated honestly.
• Is there a laboratory report? The useful ones list moisture, HMF, diastase activity and sugar profile against the Codex limits, name the laboratory and the batch, and are recent. A certificate with no batch number is decoration.
• Does the price make sense? The JRC's finding that authentic honey costs several times more than syrup is a reminder that a very cheap honey has to save money somewhere.
• What does the seller say about its limits? A seller who claims that honey cures disease is telling you something about their reliability, and none of the research in this article supports such claims.

Three common shopping situations

The supermarket squeezy bottle. Clear, uniformly golden and still liquid after a long shelf life, this kind of honey has often been warmed and finely filtered, and it may be a blend of harvests or origins. That is not automatically a problem. What you should look for is a stated country or countries of origin and whether the label says it is a blend. The sugar content is the same as any other honey.

The farmers' market jar. Cloudy, possibly setting soft, and sold by the person who keeps the bees. You can ask about location, harvest year and whether it was heated, and a good beekeeper will answer easily. Direct purchase gives you traceability that no label can, though it does not change the sugar dose.

The online jar with a premium origin story. A mountain, forest or exotic-flower claim at a higher price is only worth paying for if the seller can back it with batch information. The more a listing leans on health claims rather than origin and handling, the less weight the marketing deserves.

What the research cannot tell us yet

It helps to be plain about the limits of what we know, because they cut both ways: against exaggerated claims for raw honey, and against the idea that processing makes honey worthless.

• We found no trial that compares raw and processed honey of the same origin head to head on health outcomes in people. The raw honey subgroup in the Toronto meta-analysis compares different trials, floral sources and countries.
• The Polish study compared raw local honeys of single floral types with imported blends, so origin, floral source, blending and age are all tangled together with heating.
• Antioxidant assays such as FRAP and total phenolic content describe how a sample behaves chemically in the laboratory. They do not measure what happens in the body.
• Diastase is used as a marker of freshness and gentle handling. The research we reviewed does not show that the enzyme activity in a spoonful of honey has an effect on human health.
• The EU fraud figures describe suspicious screening results, and follow-up investigations were still needed to confirm individual cases.

Storage, crystallisation and gentle warming

Honey keeps for a long time in a sealed jar at room temperature away from heat and moisture. It crystallises with time, because glucose separates out of the water, and crystallised honey is normal, not spoiled. Raw honeys with a high glucose content set quickly, and processed honeys are often heated precisely to delay this. The main driver is the balance of the two principal sugars. In the three Mexican honeys measured in the 2023 Foods study, fructose ranged from 272 to 395 g per kg and glucose from 232 to 356 g per kg, giving fructose to glucose ratios of 1.00, 1.17 and 1.45. Glucose is the less soluble of the two, so honeys with proportionally more glucose generally crystallise sooner, while fructose-rich honeys tend to stay liquid for longer. That is a chemical property of the honey, not a sign of purity or adulteration in either direction. If you want liquid honey again, stand the closed jar in warm, not hot, water and stir. That uses far less heat than industrial liquefying and, on the evidence of the temperature studies above, leaves the markers largely intact. Avoid microwaving honey, which can overheat parts of the jar. Signs that honey has taken up moisture and started to ferment, such as foam or a sour smell, correspond to the moisture cap in the Codex standard and mean the honey should be discarded.

Who should not eat honey at all

Children under 12 months should never be given honey, raw or processed, in any form. Honey can contain spores of Clostridium botulinum, which cannot be controlled in an infant's immature gut and can cause infant botulism, and cooking does not destroy the spores. Raw honey is no more and no less risky than processed honey in this respect. Our babies and honey explainer covers the mechanism, and the risk is the reason an infant should not be given even a taste.

Origin claims: mountain honey and forest honey

Words like mountain, forest or wildflower describe the landscape where the bees foraged. They can point to real differences in flavour and floral source, but they are not quality measures. An earlier version of this page generalised that high-altitude temperatures concentrate nectar sugars and preserve enzyme quality. We could not source that as a general rule, so we have removed it. What we can say honestly is that where a honey was made is worth knowing and worth asking about, and that any origin claim is only as good as the traceability behind it.

Where our honey fits

We sell a Raw Organic Honey in a 200 g jar. We are describing what the labels on our product mean in the same terms used in this article: raw refers to how the honey is handled, not to any health effect, and the sugar content is the same as any other honey, so the limits in our dosage guide apply. If you would like to know more about origin or handling of a jar before you buy it, ask us, and treat any seller who cannot answer as a warning.

The bottom line

Raw and processed honey are not the same, and the laboratory can tell them apart: heat and long storage lower enzyme activity and raise HMF, and the best available comparison found markedly higher enzyme, protein and antioxidant markers in raw local honeys than in imported blends. But the same evidence shows that "raw" is an unregulated word, that short gentle warming changes little, that raw and processed honey have the same sugar and calories, and that there is no proof raw honey is better for your blood sugar or your heart. Choose raw honey if you value minimal handling and flavour, ask where it comes from, and treat the sugar with the same respect as any sweetener. And never give it to a baby.

Frequently asked questions

What is the real difference between raw and regular honey? Raw honey is honey that has been extracted and strained without added heat. Regular commercial honey is often warmed and finely filtered. The measurable differences are in heat-sensitive markers such as diastase activity and HMF, and in pollen content. Calories and sugar are essentially the same.

Does pasteurised honey lose all its nutrients? No. Honey is mostly sugar and water, and heat does not change that. Temperature experiments show that short heating to 50 to 80 degrees Celsius reduces diastase activity and slightly raises HMF, but it does not significantly degrade overall quality. What suffers most is the small population of enzymes and heat-sensitive compounds.

Is raw honey safer than processed honey? No. Raw honey is not sterile, and the botulism risk for infants is identical in raw and processed honey. Neither is safer for a baby, and for adults both are safe in normal amounts.

Can people with diabetes eat honey? Honey is a carbohydrate and it counts in a meal plan like sugar. Individual honeys have glycemic indexes that range from medium to high in the studies we found, and one review suggested high intakes may worsen glucose in type 2 diabetes. Agree an amount with your clinician or dietitian.

Is crystallised honey spoiled? No. Crystallisation is normal and does not mean the honey has gone off. Warm the closed jar gently in warm water if you prefer it liquid.

How can I tell if honey is real at home? You cannot reliably. The European Commission's Joint Research Centre noted that a simple test is not sufficient to detect adulteration. Rely on a traceable source, an honest label and, where possible, a batch-specific laboratory report.

Should I refrigerate honey? There is no need. A sealed jar at room temperature is fine. Freezing at minus 18 degrees Celsius preserved quality markers in the 2022 storage study, but it is not necessary for ordinary use.

How much raw honey should I eat a day? The same as any honey, because the sugar is the same. Our honey dosage guide turns the WHO, AHA and US guidelines into spoons.

How we researched this article

Standards come from the Codex Alimentarius Standard for Honey and the National Honey Board. Laboratory findings come from the peer-reviewed studies listed below, retrieved through PubMed, and fraud findings come from the European Commission's Joint Research Centre. Ranges quoted from studies are as reported by the authors, and we point out where a study is small or mixed. Where an earlier version of this page relied on figures we could not verify, we replaced them. This article is general information, not medical advice, and it was last reviewed on 21 September 2026.

References

1. Codex Alimentarius. Standard for Honey, CXS 12-1981 (adopted 1981, revised 1987 and 2001, amended 2019 and 2022). fao.org
2. Milek M, Bocian A, Kleczynska E, Sowa P, Dzugan M. The comparison of physicochemical parameters, antioxidant activity and proteins for the raw local Polish honeys and imported honey blends. Molecules 2021;26(9):2423. doi:10.3390/molecules26092423
3. Characterisation of viscosity, colour, 5-hydroxymethylfurfural content and diastase activity in raw rape honey (Brassica napus) at different temperatures. Journal of Food Science and Technology 2016. PubMed 27413239
4. Finding the optimum treatment procedure to delay honey crystallization without reducing its quality. Food Chemistry 2022. PubMed 35124485
5. Ahmed A, Tul-Noor Z, Lee D, et al. Effect of honey on cardiometabolic risk factors: a systematic review and meta-analysis. Nutrition Reviews 2023;81(7):758-774. doi:10.1093/nutrit/nuac086
6. Akhbari M, Jabbari M, Ayati MH, Namazi N. The effects of oral consumption of honey on key metabolic profiles in adult patients with type 2 diabetes mellitus and nondiabetic individuals: a systematic review of clinical trials. Evidence-Based Complementary and Alternative Medicine 2021;2021:6666832. doi:10.1155/2021/6666832
7. Palomo-de Leon BA, Castro H, Sanchez-Murillo ME, et al. Glycemic and satiety response to three Mexican honey varieties. Foods 2023;12(19):3670. doi:10.3390/foods12193670
8. Arcot J, Brand-Miller J. A preliminary assessment of the glycemic index of honey. Report for the Rural Industries Research and Development Corporation, Australia, 2005. agrifutures.com.au
9. European Commission, Joint Research Centre. Food fraud: how genuine is your honey? (23 March 2023), reporting the coordinated action "From the Hives". joint-research-centre.ec.europa.eu
10. National Honey Board. Honey FAQ (description of raw honey). honey.com
11. California Department of Public Health, Infant Botulism Treatment and Prevention Program. Prevention. infantbotulism.org
12. Infant botulism. StatPearls, NCBI Bookshelf. NBK493178