Flour by cup vs by gram — the 33% variance that breaks baking
Scoop-and-shake fills a cup with 160 g of flour. Spoon-and-level fills the same cup with 120 g. Same measuring cup, same flour, 33% more mass in one — the single largest source of baking failures that get blamed on 'bad recipes' or 'the oven'.
Flour is compressible. Its bulk density depends on how much air is trapped between particles, which depends entirely on technique. This is why every serious baking book published since 2010 quotes gram weights: it removes the technique variable. King Arthur's standard is 120 g per cup for all-purpose flour, but 'the internet' averages 140 g because scoop-and-shake is more common than spoon-and-level.
Reference weights per US cup
| Flour | Spooned (g) | Scooped (g) | Notes | |---|---|---|---| | All-purpose (AP) | 120 | 160 | King Arthur standard 120 | | Bread flour | 120 | 165 | Slightly denser, higher protein | | Cake flour | 114 | 155 | Finer particle, packs tighter | | Pastry flour | 113 | 150 | Between AP and cake | | Whole wheat | 128 | 170 | Bran adds mass | | Whole wheat pastry | 120 | 160 | Finer grind of whole wheat | | Rye | 102 | 145 | Very low gluten, loose pack | | Semolina | 168 | 200 | Coarser, denser | | Almond flour | 96 | 120 | Fat content lowers density | | Coconut flour | 112 | 150 | Absorbs water aggressively | | Rice flour | 158 | 190 | Fine rice, packs dense | | Buckwheat | 120 | 155 | Similar to AP | | Chickpea (besan) | 92 | 130 | Very light | | GF 1:1 blend | 140 | 170 | Bob's Red Mill, King Arthur |
Why volume fails specifically for flour
Wheat kernels are milled into particles of ~10–100 μm. Loose-packed, they trap ~40% air by volume. Compacted (by scoop-and-shake, by tapping the cup, by settling during shipping), the air drops to ~25%. Same cup volume, different mass. Sugar is crystalline and does not compress this way — a cup of granulated sugar is 200 g whether you spoon or scoop. Flour is uniquely bad.
Worked example — the failing cookie recipe
A chocolate chip cookie recipe calling for '2 cups AP flour' at the King Arthur standard is 240 g. Same recipe with scoop-and-shake gives 320 g — 33% more. The cookies:
- **240 g flour (correct):** spread wide, thin, crisp edges, chewy centre.
- **320 g flour (over-scooped):** stay in tight balls, cakey, pale, dry.
This is the number one 'why did my cookies fail' cause. Weigh the flour.
Spoon-and-level, correctly
The King Arthur method: fluff the flour in the container with a whisk or fork. Spoon it into the measuring cup lightly — do not tap. Level off with a straight edge (back of a knife, scraper). Result: 120 g ± 3 g reliably. Total time: 15 seconds per cup, once you have the habit.
Weight is faster than volume anyway
The stopwatch test: weighing 500 g of flour on a $15 kitchen scale takes 8 seconds (tare bowl, pour until display reads 500). Cup-measuring 500 g takes ~35 seconds (four cups spooned-and-levelled, one measured, more cleaning). Weight is faster and always accurate. The only reason cups persist in US home recipes is inertia.
Baker's percentage — flour as reference
Professional bakeries use baker's percentage: flour = 100%, every other ingredient expressed as % of flour weight. A '70% hydration' ciabatta is 700 g water per 1000 g flour. A '2% salt' bread is 20 g salt per 1000 g flour. This scaling works flawlessly — but only if flour is measured by weight.
When cups are safe
Small volumes of liquids (water, milk, oil) — a 5% cup error is invisible. Chopped vegetables, herbs, nuts — cup measurements are already approximations. Whole spices ground to taste. Everything else in a baked good: weigh.
The special case of cake flour
Cake flour is often 'sifted, then measured' in older recipes; modern recipes say 'measure, then sift'. The difference is 15 g per cup (100 sifted vs 115 unsifted). If the recipe does not specify, and it is from before 2000, assume sifted-first (lighter). If after 2010, measure-first is the norm.
The five-second recap
1 cup AP flour = 120 g if spooned, 160 g if scooped. Weigh in grams; it is faster and always right. If a recipe gives only cups, treat 120 g as the intended value unless the author explicitly says 'scoop-and-shake'.
Try the converters mentioned in this article
FAQ
What is one cup of AP flour in grams?+
120 g by the King Arthur / America's Test Kitchen standard (spooned and levelled). Older recipes and internet averages skew higher (140 g); scoop-and-shake reaches 160 g.
Does the flour brand matter?+
A little. Gold Medal AP is ~127 g/cup spooned. King Arthur is 120 g. Bob's Red Mill 1:1 GF is 140 g. Differences under 10% within a category; keep to one brand for consistency.
Can I convert an old cup-based recipe to grams once and reuse?+
Yes. Convert once using 120 g per cup for wheat flours, then annotate the recipe. All future bakes are repeatable — no more 'this recipe used to work'.
Why do European recipes use grams by default?+
European kitchens have kept scales as standard equipment; US kitchens phased them out mid-20th century in favour of nested measuring cups. Cultural, not technical.
What about volume for very small amounts?+
Under 2 tsp (10 g flour range), a scale reads unreliably at 1 g resolution. Teaspoons are fine for tiny quantities — the percentage error at absolute mass is small.
This article was written by The Turbo Unit Converter engineering desk (Culinary metrology) and last reviewed on 2026-07-10 against NIST SP 811 and the BIPM SI Brochure. Read our full editorial policy.