HoneyHQ

Why honey crystallises

Granulation is not spoilage, not a sign of adulteration, and not evidence that honey is raw. It is glucose doing what glucose does — and the rate is predictable from the sugar ratio.

Honey scienceReviewed 2026-11-02

Honey is holding more sugar than it can dissolve

Honey is a supersaturated solution: it contains far more sugar than that quantity of water could normally hold. It is metastable, meaning it will stay liquid until something gives the sugar a reason to come out of solution — and then, once started, the process continues.

The sugar that comes out is glucose. Fructose is roughly twice as soluble and stays in solution; glucose reaches saturation first and crystallises as glucose monohydrate. This is why the glucose-to-fructose ratio predicts crystallisation better than anything else, including colour.

Glucose-to-fructose ratio and typical crystallisation behaviour
HoneyTypical G:F ratioCrystallisation
Acacia≈ 0.60Very slow — often liquid after a year
Chestnut≈ 0.69Slow
Heather (ling)≈ 0.80Gels rather than granulating
Orange blossom≈ 0.85Moderate
Clover≈ 0.95Moderate, fine crystals
Sunflower≈ 0.97Very fast, coarse
Rapeseed≈ 1.03Very fast — can set in the comb
Glucose-to-fructose ratio and crystallisationAcacia honey: 0.5 to 0.7, very slow — often still liquid after a year or more. Tupelo honey: 0.58 to 0.72, very slow — often still liquid after a year or more. Chestnut honey: 0.6 to 0.8, slow — months before it starts to granulate. Forest honeydew honey: 0.7 to 0.9, very slow — often still liquid after a year or more. Heather honey: 0.7 to 0.9, moderate — granulates over a few months. Lavender honey: 0.75 to 0.97, moderate — granulates over a few months. Orange blossom honey: 0.78 to 0.95, moderate — granulates over a few months. Clover honey: 0.85 to 1.05, moderate — granulates over a few months. Sunflower honey: 0.9 to 1.1, very fast — can set in the comb before it is extracted. Rapeseed honey: 0.95 to 1.15, very fast — can set in the comb before it is extracted. Dandelion honey: 0.98 to 1.18, very fast — can set in the comb before it is extracted.tends to stay liquidtends to set fastAcaciavery slowTupelovery slowChestnutslowForest honeydewvery slowHeathermoderateLavendermoderateOrange blossommoderateClovermoderateSunflowervery fastRapeseedvery fastDandelionvery fast0.50.60.70.80.91.01.11.2Glucose ÷ fructose
Each bar is the catalogue’s range for that honey type’s glucose-to-fructose ratio, with a dot at its typical value; the word after it is the record’s crystallisation tendency. Glucose is the sugar that comes out of solution, so the more of it relative to fructose, the sooner a honey sets — below about 0.8 most stay liquid for months, above about 0.95 most set within weeks. Water content, storage temperature and the pollen, air and crystals that seed it all move the rate, which is why the bars overlap and the tendency does not follow the ratio perfectly. Ranges are HoneyHQ’s synthesis from published composition, not measurements of a jar. Sources: HoneyHQ editorial synthesis. Drawing: HoneyHQ (HEKNO Ltd), original.

What else changes the rate

  • Nucleation sites. Crystals need something to start on: pollen grains, dust, air bubbles, or existing crystals. This is exactly why ultrafiltered honey stays liquid far longer — the starting points have been removed — and why a jar that has begun to set will finish quickly.
  • Temperature. Crystallisation is fastest at around 14 °C. Warmer and the sugar stays dissolved; much colder and the honey becomes too viscous for molecules to move into position. A kitchen cupboard is often close to the worst possible temperature.
  • Water content. A higher moisture honey has more solvent and crystallises more slowly, though it is also more liable to ferment.
  • Agitation. Stirring distributes seed crystals and speeds things up. This is not a defect to avoid but a tool — it is how creamed honey is made deliberately.

What to do about it

Two honest options. The first is to eat it as it is — set honey spreads better than liquid honey and does not run off toast, which is the entire commercial argument for creamed honey.

The second is to redissolve it. Stand the jar in water at around 40-50 °C and leave it, stirring occasionally, until the crystals go. It will take a while and that is fine. What matters is not rushing: high heat destroys aroma compounds, degrades enzymes and drives up hydroxymethylfurfural, and a microwave does all three unevenly and fast. Redissolved honey will also crystallise again, usually sooner, because some crystal nuclei survive.

Sources

  • Crystallisation of honey: mechanisms and control — Peer-reviewed food-science literature · Peer-reviewed article
  • Composition of American Honeys and related analytical work — United States Department of Agriculture · Peer-reviewed article

How sources are selected and weighted is set out in the sources and evidence policy.

Read next

Tools that use this

HoneyHQ saves your cookie choice on this device. Optional analytics run only if you accept — set a preference now, or change it later on the cookies page.