One Cup of Flour Can Weigh 38% More Than Another
The same cup recipe can mean 57% or 79% hydration, depending on how you fill the cup. Baker's percentages remove that spread. Here is the conversion, with the arithmetic.
One cup of all-purpose flour weighs 120 g on the King Arthur chart [1]. One test that scooped straight from the bag got 166 g [2]. That is 38% more flour in the same cup (166 / 120 = 1.383). The water does not change when you scoop harder. So one cup recipe can give a slack dough or a stiff one, and the recipe never says which.
Baker's percentages fix this. Flour is 100%. Every other ingredient is its weight divided by the flour weight, times 100 [3]. The one-line ratio: ingredient percent = ingredient grams / flour grams x 100. Below I convert two cup recipes in five steps.
What you will have at the end
- A table that turns a cup recipe into grams and percentages.
- A hydration range for the recipe, not one false number.
- A rule to scale the recipe to any size.
- A test you can run in one afternoon.
Prerequisites and "Tested with"
You need a pen or a spreadsheet. You need a digital scale for the final test. I did not bake for this post. I did not run code or the Lab. Every number below is hand arithmetic on published weights, so you can repeat it.
The recipes are template inputs, not published recipes. I made them up to be typical. Treat them as examples. The weights per cup come from the King Arthur chart [1]:
| Ingredient | Grams |
|---|---|
| Flour, 1 cup (all-purpose or bread) | 120 |
| Granulated sugar, 1 cup | 198 |
| Butter, 1/2 cup | 113 |
| Water, 1 cup | 227 |
| Table salt, 1 tbsp | 18 |
| Instant yeast, 2 tsp | 6 |
The chart does not say how to fill the cup. It tells you to use a scale [1]. Another source says its flour figure matches spoon and sweep, and that dip and sweep gives about 4.75 to 5 oz (135 to 142 g) [4]. A forum thread reports scooped cups from 140 g to 165 g [5]. The 166 g and 138 g figures come from one blog [2], so I treat them as one data point. The forum and the other source point the same way, which is why I use them as a plausible range, not as a measured one.
Flour weights I use for "one cup": 120 g (spoon and sweep), 138 g (spoon and level), 166 g (scoop) [2].
Step 1: Write the template recipe in cups
Loaf (template): 3 cups flour, 1 1/4 cups water, 1 tsp table salt, 1 tsp instant yeast. Cake (template): 2 cups flour, 1 cup sugar, 1/2 cup butter.
Step 2: Convert every non-flour item to grams
Water: 1.25 x 227 = 283.75, so 284 g. Salt: the chart gives 18 g per tbsp, so 1 tsp is 18 / 3 = 6 g. Yeast: 6 g per 2 tsp, so 1 tsp is 3 g. Sugar: 198 g. Butter: 113 g.
Step 3: Convert flour three ways
Flour is the only item with a wide spread, so convert it three times. Multiply the cup count by each flour weight per cup: 120 g, 138 g and 166 g. The loaf uses 3 cups. The cake uses 2 cups.
Loaf flour: 3 cups x 120 = 360 g | x 138 = 414 g | x 166 = 498 g
Cake flour: 2 cups x 120 = 240 g | x 138 = 276 g | x 166 = 332 g
The loaf flour spans 138 g (498 - 360). The cake flour spans 92 g (332 - 240). The water, sugar and butter stay fixed, so each flour weight gives a different recipe.
Step 4: Divide by flour
Loaf (water 284 g, salt 6 g, yeast 3 g)
flour g water % salt % yeast %
360 78.9 1.67 0.83
414 68.6 1.45 0.72
498 57.0 1.20 0.60
Cake (sugar 198 g, butter 113 g)
flour g sugar % butter %
240 82.5 47.1
276 71.7 40.9
332 59.6 34.0
This is hand arithmetic, not Lab output. Check one cell: 284 / 360 = 0.789, so 78.9%.
Read the loaf rows. The "same" loaf spans 57.0% to 78.9% hydration, a spread of 21.9 points. That covers a stiff bagel dough up to a wet, open ciabatta-style dough. In an earlier post I argued that 75% hydration is not one thing across flours. I extend that here: before flour protein matters, the flour weight itself is uncertain by a third. The dough was 70% water and 100% honest only if someone weighed the flour.
The cake spans 59.6% to 82.5% sugar. Sugar and fat set structure in cakes, and these rows are far apart. The 82.5% row is not a typical cake and the 59.6% row is not either. A cup recipe cannot tell you which one the author baked.
Step 5: Scale with one multiplier
Pick the flour you want, then multiply everything by flour / 100. For a 500 g flour loaf at the 68.6% row:
water: 500 x 0.686 = 343 g
salt: 500 x 0.0145 = 7.3 g
yeast: 500 x 0.0072 = 3.6 g
Each row sums to more than 100%. For the 360 g loaf: 100 + 78.9 + 1.67 + 0.83 = 181.4%. That is normal [3]. The Caputo example (1000 g flour, 600 g water, 30 g salt, 3 g yeast) totals 163.3% [6]. Watch for a different convention: some European usage sums flour and water to 100% [7]. Check which one a source uses.
Dough and oven temperature
Percentages say nothing about heat. State both. For the loaf I would aim for a mixed dough near 24 C and an oven near 230 C. These are my template values, not measured ones. Water at 78.9% hydration at 24 C is a different object from the same dough at 30 C, because fermentation speed changes with temperature (see the fridge post).
How to verify it worked
- Add up the grams in the converted recipe. Divide the total by flour grams. It should equal the sum of the percentages / 100. For the 360 g loaf: 653 / 360 = 1.814. It matches.
- Rebuild the recipe from the percentages. Multiply each percent by flour. You should get the original grams back.
- Check the hydration against the dough. A dough near 57% should feel firm. A dough near 79% should be sticky. If it does not, your flour weight is wrong or the flour absorbs differently. Flour protein and absorption also shift how wet a hydration feels, which is the point of the earlier post.
When it fails
I have no program output with error messages here, so I list the failure modes by symptom.
- Symptom: the percentages add up to exactly 100%. You normalised flour and water to 100% [7]. Divide again by flour only.
- Symptom: you get 1,000% for salt. You divided flour by salt. Put flour in the denominator.
- Symptom: cup recipe gives no method. The range in Step 3 is the honest answer. Do not pick one value. Bake at the middle and adjust water by feel, or weigh and report back.
- Symptom: brand salt. One tablespoon of kosher salt is 8 g (Diamond Crystal) or 16 g (Morton's), against 18 g for table salt [1]. A "tablespoon of salt" can differ by a factor of 2.25 (18 / 8). Pick the brand first.
- Symptom: whole wheat. The chart gives 113 g per cup [1], not 120 g. A fixed cup weight does not carry across flours.
This rule has limits. Percentages do not tell you touch, which a skilled baker uses and I do not. They also assume the flour is dry and equal in absorption. For a single mixed dough at home, they beat cups every time I have tested them on paper.
Your one-afternoon test
Take any cup recipe you own. Weigh one level cup of flour three ways: scoop, spoon and level, and spoon and sweep with a fluffed bag. Write each grams value. Divide the recipe's water grams by each. If your three hydrations span more than 10 points, the cup recipe is not one recipe. I have not run this test myself, and I expect a spread of about 15 to 20 points, based on the 120 to 166 g range above. If your spread is under 5 points, tell me, because that would contradict the sources.