Why Doubling a Recipe Fails (And What to Scale Instead)
September 2, 2026
You had eight people coming instead of four, so you did the obvious thing. Two cups of flour became four. One egg became two. Everything on the list got multiplied by two, and the arithmetic was flawless — you checked it twice. Then you poured all of it into the same 9-inch round pan the recipe called for, set the timer for the same 35 minutes, and came back to a cake with a burnt rim, a sunken middle and a wet seam of raw batter under the crust.
Nothing was wrong with your maths. The problem is that you multiplied a recipe, and a recipe is not only a list of ingredients. It is also a shape and a chemistry, and neither of those obeys the multiplier.
Here is the one line worth remembering: mass scales. Geometry and chemistry do not.
The short answer, in one table
| What you’re scaling | Multiply by 2? | What to do instead |
|---|---|---|
| Flour, sugar, butter, oil, milk, water | Yes | Straight ×2. This is the easy 90%. |
| Whole eggs | Yes | ×2 is fine going up. Halving is the awkward direction. |
| Salt and assertive spices | No | Start around 1.5×, then taste and top up. |
| Baking powder / soda | Only at the same depth | Same depth ×2. Deeper batter: cut ~10–25% and drop the oven. |
| Water for rice, simmering liquid | No | Evaporation doesn’t double. Add a fraction, not a factor. |
| Pan | No | You need double the area, not double the label. |
| Oven time | No | It tracks batter depth. Check early, judge by doneness. |
| Oven temperature | Depends | Deeper batter: down 25 °F / 15 °C, longer bake. |
The first row is the reason recipe scaling feels like it should be easy — most of the list really is just multiplication. Paste your ingredients into the recipe scaler and it rewrites every quantity for you, fractions and ranges included, so you never fumble “half of ¾ cup” at the counter. What it deliberately does not touch is the pan, the clock or the seasoning, and the rest of this article is why.
The ingredients that really do just double
Start with the good news, because it is most of the recipe. Flour, sugar, butter, oil, milk, water, cocoa, chocolate, oats, nuts, most vegetables and most proteins scale straight. Doubling them changes the quantity and nothing else — the ratio between them, which is what actually defines the recipe, is untouched.
That is why bakers who think in baker’s percentages find scaling trivial: if flour is 100% and water is 68%, those percentages are identical whether you are making one loaf or forty. The multiplier disappears entirely. America’s Test Kitchen makes the same point from the other direction, under the heading “Don’t Overthink It” in their tips for doubling: “Want to double a salad, stir-fry, dressing, sauce, oven-roasted salmon, pasta, or other simple recipe? Just multiply each ingredient by two.” Nothing about those dishes has a geometry.
The trouble starts the moment the food has to sit somewhere and get hot.
Geometry: the pan is the thing that broke your cake
A pan’s capacity is its area, and area grows with the square of the width. For a round pan that is the formula our cake pan converter runs on:
area = π × (diameter ÷ 2)²
Square and rectangular pans are simply length × width. Those three formulas are the whole calculator, and they produce numbers that consistently surprise people:
| Pan | Area | Versus a 9-inch round |
|---|---|---|
| 8-inch round (20 cm) | 50 in² / 314 cm² | 0.79× |
| 9-inch round (23 cm) | 63.6 in² / 415 cm² | 1.00× |
| 10-inch round (25 cm) | 79 in² / 491 cm² | 1.23× |
| 8-inch square (20 cm) | 64 in² / 400 cm² | 1.01× |
| 9 × 13-inch (23 × 33 cm) | 117 in² / 759 cm² | 1.84× |
| 12¾-inch round (32.5 cm) | 128 in² / 830 cm² | 2.00× |
The ratios in that last column come from the inch figures. The metric sizes beside them are the near equivalents you would actually buy, not exact conversions, so they give slightly different answers — feed the converter its default 23 cm round to 20 cm round and it returns 0.76×, “make 24% less batter.” Compare like with like and the sums agree.
Read the last row again. To bake a doubled 9-inch cake in one round pan, you need a pan almost thirteen inches across. Going up to a 10-inch pan — which feels like a meaningful jump — buys you 23% more room, not 100%. The shortcut is that doubling the area means multiplying the diameter by √2, about 1.41×; halving it means multiplying by 0.71.
This is also where a widely repeated piece of advice deserves a footnote. America’s Test Kitchen suggests reaching for a 13 by 9-inch pan when you double a recipe written for an 8-inch square, and it is good advice — but compared with that 8-inch square (64 in²), the 13 by 9 is 1.83× the area, not 2×. Double the batter goes in about 9% deeper than the original. (The table above measures the same 13 by 9 against a 9-inch round, which is very slightly smaller at 63.6 in², so it comes out at 1.84× — different baseline, not a contradiction.) That is usually fine. It is not nothing, and it is exactly the sort of margin that turns into an extra ten minutes in the oven.
The clean answer, and the one professional bakers give, is to use two pans. Two 9-inch rounds are exactly 2.00× one 9-inch round, at identical depth, which means the original time and temperature still apply. ATK put it plainly: you can mix a double batch of dinner rolls by doubling the ingredients, “but you should still bake each batch in its own pan.”
One more constraint worth knowing before you pour: pan maker Fat Daddio’s says to fill a cake pan between one-half and two-thirds full — “any less and the cake will be flat. Any more and it will overflow or dome aggressively.” A doubled batter in a single pan blows straight through that ceiling. The burnt ring on your oven floor was predictable from the geometry alone.
Time: it tracks depth, not total volume
Here is the misconception that ruins more doubled bakes than any other: people assume twice the batter needs roughly twice the time, or at least some extra time proportional to the amount. It doesn’t work that way at all.
ATK states the actual rule in no, you can’t always just double a recipe: bake time and temperature are determined by “the depth and surface area of the dough or batter you’re working with.” Heat travels inward from the pan walls and the surface. If you split a doubled batter across two pans, the depth is unchanged, every point in the batter is the same distance from the heat as before, and the bake time barely moves. If you pour it all into one pan, the depth doubles, the centre is twice as far from any hot surface, and the time goes up sharply while the edges keep cooking at the old rate. Burnt rim, raw middle. That is the whole failure, in one sentence.
Fat Daddio’s publish numbers for exactly this, framed as pan depth rather than as doubling, which is the same physics:
| Change in batter depth | Oven temperature | Bake time |
|---|---|---|
| 2-inch → 3-inch pan | −25 °F (−15 °C) | +15 to 25 minutes |
| 3-inch → 4-inch pan | −15 to 25 °F | +20 to 40 minutes |
| 2-inch → 4-inch pan | −25 to 50 °F | +30 to 60 minutes or more |
The University of Wyoming Extension gives the same instruction as a rule of thumb in its recipe-scaling guide: “for thicker baked goods, back the temperature down to about 25°F and cook for a little longer.” Lowering the oven is not a workaround, it is the point — a cooler oven gives the centre time to catch up before the outside sets hard. If you need to move between scales, the oven temperature converter will hand you the matching °C or gas mark for whatever you land on.
And then stop trusting the clock entirely. ATK’s advice for any scaled recipe is to lean on the recipe’s doneness cue rather than its timing — “an inserted toothpick coming out clean,” the colour, the spring-back. Fat Daddio’s go further and suggest an instant-read thermometer, aiming for an internal 200–210 °F (93–99 °C) in a finished cake; the WebstaurantStore cake troubleshooting guide lands on 205 °F. A doubled cake is a new recipe as far as the oven is concerned, and its timing has to be measured, not inherited.
Chemistry, part one: salt and spices
Perception of seasoning is not linear. Double the salt in a doubled dish and it very often reads as saltier, not the same — and unlike a pan, you cannot undo it.
How much less should you use? This is where honest sources stop agreeing, and it is worth saying so rather than inventing a number. The University of Wyoming Extension commits to a figure: “you’ll likely need to increase the seasonings by an extra 50% in recipes that are doubled” — that is 1.5×, not 2×. America’s Test Kitchen, meanwhile, declines to give a multiplier at all in either of its doubling articles; its guidance throughout is to judge the finished dish rather than to trust a factor.
That spread is not sloppiness. The right factor genuinely depends on the dish. Salt in a brine or a bread dough is a ratio and scales dead straight — a 2% dough is 2% at any size. Salt as final seasoning in a stew is a perception and does not. Ground spices sit in between, and volatile ones like cayenne, cloves and black pepper punch above their weight in a big pot because there is more time and more surface for them to bloom.
The practical method, which no source disagrees with:
- Scale the salt and strong spices at about 1.5× when doubling — roughly three quarters of what the multiplier says.
- Cook it.
- Taste, and add the rest in small increments.
You can only add. That asymmetry is the entire argument. In the recipe scaler you can untick individual lines so salt, spices and leavening stay at their original amounts while everything else doubles, then dial them up by hand — which is exactly this workflow.
Chemistry, part two: leavening
You will read everywhere that you must cut the baking powder when you double a recipe. That advice is half right, and the half that is missing is the important one.
Baking powder is dosed as a ratio to flour, not as a fixed amount. Wikipedia’s baking powder entry gives the standard rate: “one teaspoon (5 g) of baking powder is used to raise a mixture of one cup (120 g) of flour, one cup of liquid, and one egg.” Double the flour and you genuinely do need double the gas to lift it. If you split the doubled batter across two pans at the original depth, double the leavening is correct, and cutting it will give you a dense cake for no reason.
The advice to reduce it exists because most people doubling a recipe do not keep the depth the same. In a deeper batter, the same proportion of gas has to lift a taller column before the structure sets, the bake runs longer, and the bubbles have more time to coalesce. Colorado State University Extension describes this failure mode precisely in its high-elevation guide — excessive rising “stretches the cell structure of the cake, making the texture coarse, or breaks the cells, causing the cake to fall.” The trigger there is low air pressure rather than pan depth, but the mechanism is identical: gas expanding faster than the crumb can set around it. WebstaurantStore describe the milder version, where “an excess of baking powder can cause the middle of your cake to rise too quickly and hold that dome shape.” There is a flavour cost too — Wikipedia notes that unreacted acids from excess powder “often lend an unpleasant taste to food,” and that some formulations turn outright bitter.
So the honest rule is conditional, and no source I could find puts a defensible single number on it:
- Same depth, two pans: scale leavening straight, ×2.
- Noticeably deeper batter in one pan: trim the leavening somewhat — a working range is 10% to 25% off the calculated amount — and lower the oven 25 °F / 15 °C so the crumb sets before the gas outruns it. Treat that as a starting point to test, not a law, and write down what worked.
Be aware that even the size of the trim is contested. Imperial Sugar’s rule, quoted by Mashed, is to decrease the leavening by “⅛ teaspoon for every teaspoon the recipe requires” — so a recipe with 1 teaspoon of soda gets 1⅞ teaspoons in the double batch, not 2. That is a cut of about 6%, well below the range above rather than a confirmation of it. None of these sources reports a controlled test, which is the real reason the numbers wander: the right trim depends on how much deeper your batter actually got. Start small, and let the first bake tell you.
The pot problems: rice, beans and browning
Scaling failures are not just a baking phenomenon. Anything that depends on evaporation or on contact with a hot surface breaks the same way, because a pot’s surface area does not double when its contents do.
ATK found this experimentally with rice: “when doubling a batch of rice using the same conditions as we’d used for a single batch, the same quantity of water evaporated.” The absorption is linear, the evaporation is not, so doubling both rice and water leaves an excess and gives you mush. Beans have the mirror problem — twice the liquid over the same surface simply needs longer to reduce. Braises and sauces run thinner. And a doubled quantity of meat crowded into the same pan steams instead of browning, because the water coming off it has nowhere to go.
The fixes are all about restoring surface area and headroom: a wider pot (ATK’s advice for a doubled chili is to swap the Dutch oven for a stockpot, or use two), browning in batches, and adding well under double the liquid, then topping up if it is needed. Reduce first, season last.
Halving is harder than doubling
Scaling down looks safer and is fiddlier, because small numbers hit the floor of what a kitchen can measure.
- Eggs. A US Large egg is 56.7 g in the shell, which is about 50 g of actual egg once cracked. Half an egg is not a guess: beat one whole egg, weigh out 25 g, and discard the rest. A scale turns the single most annoying part of halving into ten seconds of work.
- Leavening. Half of ¾ teaspoon is ⅜ of a teaspoon, and nobody owns that spoon. Use ¼ plus a scant ⅛, or weigh it — a teaspoon of baking powder is about 5 g.
- Pans. Half the area means 0.71× the diameter, so a halved 9-inch round wants a pan a little over 6 inches (16 cm) across, not a 4½-inch one. Alternatively keep the 9-inch pan, accept a shallower cake, and start checking it 10–15 minutes early — that direction is far more forgiving than going deeper, because a thinner layer only ever bakes faster.
- Seasoning. Halving salt is the one case where scaling straight is safe, since you can always add more at the table.
The takeaway
The recipe is the ratios. Everything else — the pan, the oven, the clock, the last pinch of salt — is a separate decision that the multiplier never touched.
So do the arithmetic once, properly, and then think about the shape. Paste your ingredients into the recipe scaler, set a factor or just tell it you are going from four servings to eight, and let it rewrite the quantities — including the fractions — while you untick the lines you want to season by hand. Then work out the pan with the cake pan converter, which compares the two areas and tells you the multiplier your pans can actually support. If you bake often enough that scaling is a weekly problem, the baker’s percentage calculator removes it permanently — a recipe expressed as percentages of flour has no size at all. And when you end up dropping the oven for a deeper bake, the oven temperature converter has the °F, °C and gas mark equivalents, with the reasoning in our oven temperature guide.
Then check the middle with a thermometer, not a timer. Everything else about a doubled recipe you can fix; a raw centre you cannot.
Always follow current food-safety guidance — see our disclaimer.