Can You Substitute Honey for Sugar in Baking? The Real Limit
You can substitute honey for sugar in baking, and in most cakes and cookies the safe ceiling is around half the sugar by weight. Beyond that the crumb turns dense and slightly gummy. The cause is not the extra water everyone talks about. Honey’s free glucose and fructose let the starch set roughly three to four degrees Celsius earlier than table sugar does, so the crumb locks before it has finished rising.
How confident: the starch mechanism is measured and published (Woodbury and colleagues, Journal of Food Science, 2023). The half-and-no-further ceiling comes from extension-service test kitchens going back to at least 1987 and matches what I would expect at the bench. Nobody has published a proper dose-response curve for liquid honey in a standard butter cake, so treat the exact ceiling as a working number.
Most baking guides agree that honey can replace sugar. They do not agree on how much, and that disagreement is what sent me digging. I read every page ranking for this question, and not one of them disagreed about the method while every one of them disagreed about the number. Half a cup of honey per cup of sugar. Two thirds. Three quarters. One to one. Same four adjustments underneath, four different answers on top, and none of them explaining why.
That gap is usually a sign the question has never actually been settled in public. So I went back to the food science research, compared it against what happens on the bench, and the answer turned out to be more interesting than the ratio.
Where This Belief Came From
The familiar ruleset (use less honey, cut a quarter cup of liquid, add a little baking soda, drop the oven 25°F) traces back to National Honey Board guidance that spread through American cooperative extension bulletins from the 1980s onward. The original version carried a limit. The internet copy dropped it.
You can watch it happen in the documents. A Michigan State University Extension bulletin from June 1987 lists the same four adjustments almost word for word, and then adds a sentence that has quietly vanished from the modern versions: replacing more than half the sugar in a cake recipe with honey will give you a very heavy, dense cake.
The same four rules are still sitting on extension pages today, at Missouri Extension and in the Texas A&M nutrition sheet, and the beekeeping associations pass along the National Honey Board version too. The rules themselves are fine. What got lost in forty years of retelling is that they were written as a partial-substitution guide, not a licence to swap the whole lot.
So the honest history is not that anyone lied. It is that a careful piece of institutional advice got flattened, one blog post at a time, until only the ratio survived. Forty years of retelling, and the one sentence that actually answered the question is the one that got cut!
What Happens When You Substitute Honey for Sugar in Baking
Sugar in a cake is a structural ingredient, not only a sweetener. It holds starch granules shut until the batter has finished rising. Honey’s sugars do that job less well, so the crumb sets earlier and lower. The water and the browning are real effects too, but they are secondary.
Table sugar is sucrose, one glucose bonded to one fructose. Honey has already been taken apart: roughly 38% fructose and 31% glucose floating free, plus about 17% water and a scatter of minor sugars and acids.
That difference matters more than it sounds. Sugar dissolved in the batter competes with starch for water and raises the temperature at which starch granules swell and gelatinise. The batter therefore stays fluid for longer in the oven, the bubbles keep expanding, and the crumb only locks once it has risen. This is why a low-sugar cake comes out tight and a properly sugared one comes out open.
Here is the measurement that explains the whole question. Woodbury and colleagues at Purdue put wheat starch through differential scanning calorimetry in different sugar solutions and published the results in the Journal of Food Science in 2023. At 30% concentration, starch in sucrose set at 72.23°C. Starch in a 50:50 glucose and fructose mixture, which is close to what honey delivers, set at 68.62°C.
Three and a half degrees sounds like nothing. In a cake it is the difference between a crumb that sets after the last of the oven spring and one that sets during it. Replace a little sugar and you will not notice. Replace most of it and you get exactly the heavy, dense cake that 1987 bulletin warned about.
Three other things happen at the same time, and they are worth knowing because they are the ones you can adjust for directly:
- Faster browning. Free glucose and fructose are reducing sugars and go straight into the Maillard reaction. Sucrose has to split first. That is why the 25°F rule is real and not folklore.
- Added water. A cup of honey weighs about 340 g and carries roughly 58 g of water, which is almost exactly a quarter cup. The arithmetic behind the old rule checks out.
- Acidity. Honey runs acidic, generally somewhere between pH 3.4 and 6.1, so a pinch of baking soda gives you back some lift and tempers the tang.
And one thing you cannot adjust for: creaming. If your recipe depends on beating crystalline sugar into butter to punch air pockets into the fat, honey has no crystals to do it with. No amount of ratio maths fixes that.
What the Evidence Says
The starch-set finding is instrumented and recent. The sweetness claim underneath every substitution chart is much weaker than it looks. The staling claim points in two directions at once depending on which kind of staling you mean.
| Claim | What was measured | What they found | Strength |
|---|---|---|---|
| Honey’s sugars set starch earlier than table sugar | Wheat starch by DSC, 0 to 60% sugar solutions | At 30%, sucrose 72.23°C vs glucose-fructose 68.62°C | Strong. Peer reviewed, instrumented, 2023 |
| Honey is much sweeter, so use less | Sensory panels, fructose vs sucrose, water and cake | 1.6 to 1.9 times sweeter in dilute water. No detectable difference in vanilla cake | Moderate. Peer reviewed sensory work, but from 1979 |
| Honey keeps baked goods fresher | Amylopectin retrogradation in starch solutions | Fructose and the glucose-fructose mix sped up retrogradation. Sucrose and glucose slowed it | Strong method, weak transfer. A starch model is not a cake |
| Honey keeps baked goods fresher (the other result) | Bread loaves with honey powder replacing sucrose | Highest loaf volume, least freezable water, slowest firming | Moderate. Single study, 2013, and powder is not liquid honey |
| Half is the ceiling in cakes | Extension test-kitchen work behind the 1987 bulletin | More than half gives a very heavy, dense cake | Moderate. Institutional, but the method was never published |
| All-honey pound cake fails | One baker’s side-by-side pound cakes, 2025 | Partial collapse and a gummy layer at the base | Weak on its own. One kitchen, one recipe, but it matches the mechanism |
Read together, the picture is consistent. Every result that points at texture points the same way, and the one that seems to contradict it (the bread study) is measuring a bake where sugar was never doing structural work in the first place.
Why People Still Disagree About the Ratio
The ratio camps are not arguing about honey. They are answering different questions. One camp is matching sweetness, one is matching mass, and one is baking bread where none of it matters much.
| The case for using less honey | The case for a straight swap |
|---|---|
| Fructose measures sweeter than sucrose in solution, so gram for gram honey should over-sweeten. Cut it to two thirds and you land where the recipe intended. | Cutting the sweetener also cuts the tenderiser and the structure. You fix perceived sweetness and break the crumb, which is the more expensive mistake. |
| Honey carries water that sugar does not, so a smaller quantity keeps hydration closer to the original. | Correct the water directly instead. Then you keep the full sweetener load and still land on the right batter. |
| Long practice. Extension guidance, cookbooks and bakers have used two thirds for decades and it works. | Yeast bakers swap one to one all the time with good results, and they are not wrong either. |
The disagreement survives because the sweetness premise was measured in the wrong place. In 1979 Cardello and colleagues found fructose to be 1.6 to 1.9 times sweeter than sucrose in dilute water, and then found no detectable sweetness difference at all when each sugar was used in vanilla cake. At cake concentrations, in the presence of fat and acid, the advantage collapses. Honey also carries plenty of glucose, which is less sweet than sucrose, not more.
So the reason to use less honey in a cake is not really sweetness. It is the water and the structure. And there is a small arithmetic slip riding along with the old rule: a quarter cup of liquid is the correction for a full cup of honey. If you are using three quarters of a cup, the honey only brought about three tablespoons of water with it, so cutting the full quarter cup leaves your batter drier than you meant.
The bread camp is right too, for a reason that has nothing to do with the argument. Yeast bread does not lean on sugar for aeration or for starch-set timing. Swap the lot and nothing structural is at stake.
Test It in Your Own Kitchen
One muffin batter, four batches, a kitchen scale and a ruler will settle this for your oven in an afternoon. Measure height and press the crumb. If the ceiling is real, you will see it appear somewhere between the 50% and 100% batches.
What you need: a plain muffin or quick-bread batter you already trust, a digital scale that reads grams, mild honey (clover or acacia, not buckwheat), one muffin tin, a ruler.
- Weigh the sugar the recipe calls for. Make four batches replacing 0%, 25%, 50% and 100% of it, gram for gram, with honey.
- For every 100 g of honey you add, remove 17 g of other liquid. Nothing else changes.
- Bake all four at the same reduced temperature, 15°C or 25°F below the recipe, in the same tin, same rack, same session. Browning is not what you are testing, so take it out of the equation.
- Cool completely. Cut each one down the middle, stand the ruler against it and record the height. Photograph the cut faces together.
- Press the crumb with a thumb and note whether it springs back or stays compressed. Repeat the press test at 24 and 48 hours.
What the result means: if height falls off sharply between 50% and 100% and the crumb at 100% stays compressed, the starch-set mechanism is showing up in your kitchen. If all four are within a few millimetres, your recipe was not leaning on sugar for structure and you can swap freely in it.
What will make your test misleading: measuring honey by volume instead of weight, using a dark honey whose flavour swamps your judgement, baking the batches in separate sessions, judging by eye instead of a ruler, or tasting them at different temperatures. And one muffin per batch is not a result. Bake three of each and compare the middle one.
What This Means for Your Baking
Work in grams, cap the swap by bake type, and correct the water rather than shrinking the sweetener. Those three moves cover almost every recipe you will want to convert.
- Use this conversion. It is the one I would work from, because it corrects the water directly instead of shrinking the sweetener. For every 100 g of sugar you replace, use 100 g of honey and remove 17 g of other liquid. Drop the oven 15°C or 25°F and start checking ten minutes early. Add a quarter teaspoon of baking soda per cup of honey if the recipe has none.
- Cap it by what the sugar was doing. Yeast breads and quick breads take a full swap. Butter cakes and muffins stop at about half. Creamed cookies stop lower, around a quarter to a third, because you are also losing the creaming. Meringues, macarons, buttercream, sponge and shortbread are not candidates at all.
- Start at 50% even where more is possible. You get most of the flavour and moisture for none of the structural risk, which is what that 1987 bulletin was quietly telling everyone.
- Pick a mild honey for anything delicate. Clover or acacia for citrus and vanilla bakes. Save buckwheat for gingerbread, where you want it to shout.
- One safety line that is not negotiable: never feed honey, or anything baked with it, to an infant under twelve months. Honey can carry Clostridium botulinum spores and infant botulism is a real risk. Baking temperatures do not reliably destroy the spores.
And do not swap honey in expecting it to be a health upgrade. Per spoonful it carries slightly more sugar and slightly more calories than granulated sugar. Use it because you like what it does to the flavour, which is a perfectly good reason.
What We Still Don’t Know
Three things stay genuinely open, and I would rather say so than round them off.
Nobody has published a proper dose-response study on liquid honey in a standard butter cake. The half-and-no-further figure comes from extension test kitchens whose method was never written up, and the sharpest instrumented work was done on starch solutions rather than real batters full of fat and egg. Both point the same way, which is reassuring, but the ceiling is an inherited number and not a measured one.
The staling question is unresolved in an interesting way. Fructose sped up starch retrogradation in the 2023 solution work while honey clearly does hold moisture in a finished loaf. Those are two different kinds of going stale, and I have not found anyone who has separated them properly in one cake over one week.
And varietal effects are unmeasured. Honeys range from roughly 31% to 44% fructose depending on the flower. Whether that spread is wide enough to shift the starch-set temperature enough to notice in a home oven is, as far as I can tell, an open question.
More Baking Science from the Lab
- Molten Nutella brownies and the frozen-core trick, where sugar is doing structural work you can taste.
- Ten croissant problems and how to fix them, a longer troubleshooting walk through laminated dough.
- Berry brunch bake, a forgiving batter if you want somewhere low-stakes to try a 50% honey swap.
- Our healthiest recipes, for the bakes where sweetener choices matter most.
Frequently Asked Questions
How much honey replaces 1 cup of sugar?
Work in grams rather than cups. Replace up to half the sugar weight with the same weight of honey, then remove 17 g of other liquid for every 100 g of honey you added. In cup terms that lands near three quarters of a cup of honey for a full cup of sugar, but the gram version is the one that survives being scaled.
Does honey make cakes dense?
Past roughly half the sugar, yes. Honey’s free glucose and fructose let starch gelatinise about three to four degrees Celsius earlier than sucrose does, so the crumb sets before the cake has finished rising. Below half the effect is usually too small to see.
Why do you lower the oven temperature when baking with honey?
Honey’s glucose and fructose are reducing sugars that enter the Maillard reaction immediately, while sucrose has to break apart first. The surface therefore browns and can burn well before the centre is set. Dropping the oven 15°C or 25°F buys the middle enough time.
Can I substitute honey for sugar in cookies?
Only partly. Around a quarter to a third works. Creamed cookie doughs use sugar crystals to cut air into the butter, and honey has no crystals to do that with, so a bigger swap gives you a flatter, softer, chewier cookie whether you wanted one or not.
Is honey healthier than sugar in baking?
Not meaningfully. Honey carries slightly more calories and carbohydrate per spoonful than granulated sugar and behaves as sugar in the body. Choose it for flavour and moisture, not as a health swap, and never give it to a child under one year old.
Sources
- Woodbury, T. J., Pitts, S. L., Pilch, A. M., et al. Mechanisms of the different effects of sucrose, glucose, fructose, and a glucose-fructose mixture on wheat starch gelatinization, pasting, and retrogradation. Journal of Food Science, 2023. ift.onlinelibrary.wiley.com
- Cardello, A. V., et al. Relative sweetness of fructose and sucrose in model solutions, lemon beverages and white cake. Journal of Food Science, 1979. ift.onlinelibrary.wiley.com
- Michigan State University Cooperative Extension. Honey, Extension Bulletin E-2071. 1987. archive.lib.msu.edu
- University of Missouri Extension. Using and Storing Honey, GH1120. extension.missouri.edu
- Texas A&M AgriLife, Food & Nutrition Unit. Honey vs Sugar. nutrition.tamu.edu
- North Carolina State Beekeepers Association. Honey FAQ, drawing on National Honey Board guidance. ncbeekeepers.org
- Effect of honey powder on bread quality. Journal of Food Processing & Technology, 2013. doi.org
- Gray, E. How sugar affects cake texture, including a side-by-side honey pound cake. Baking Sense, 2025. baking-sense.com
About Chef Arafat Hossain
Arafat Hossain is a Head Pastry Chef, food creator, and AI-powered culinary educator with more than 12 years of professional experience in patisserie, artisan bakery, and dessert development across India, the UAE, Kuwait, and Saudi Arabia. He combines classical pastry techniques with modern food science to create recipes that are practical, visually stunning, and easy to recreate at home. Through ChefArafat.ai, he shares baking tips, recipe experiments, and AI-assisted food content with a global community of dessert lovers and home bakers. Sweetener substitution sits close to his daily work: production bakery formulas live or die on how a crumb sets.