How to convert a recipe between different baking pan sizes

Scale by area, not by width — a 23 cm tin holds a third more than a 20 cm one. What the multiplier covers, and the four things it does not.

Published 2 August 2026

You’ve got a recipe for a 20 cm round tin and a 23 cm tin in the cupboard. The difference sounds small. It’s about a third more cake.

That gap, between how big a tin sounds and how much it holds, is the whole problem. It’s why cakes come out flat, why brownies come out raw in the middle, and why the same recipe works for one person and not another.

Scale by area, not by width

Batter fills a footprint. For a round tin the area is πr², so it goes as the square of the diameter:

20 cm round  →  π × 10²  ≈  314 cm²
23 cm round  →  π × 11.5² ≈  415 cm²

415 ÷ 314 = 1.32

A 15% wider tin needs 32% more batter. Scale by width instead, multiplying by 1.15, and you underfill by about an eighth. That’s the difference between a cake and a disappointingly thin one.

The same arithmetic in the other direction explains why a recipe doubled into a tin “twice as wide” overflows: double the diameter and you quadruple the area.

Some useful reference points, all relative to a 20 cm round:

Tin	Area	Multiplier
15 cm round	~177 cm²	0.56×
18 cm round	~254 cm²	0.81×
20 cm round	~314 cm²	1.00×
20 cm square	400 cm²	1.27×
23 cm round	~415 cm²	1.32×
23 cm square	529 cm²	1.68×
23 × 33 cm rectangle	759 cm²	2.42×

Look at the square-versus-round line. A 20 cm square holds about 27% more than a 20 cm round, because the round tin is the square with its corners cut off. It’s the most common accidental substitution in baking, and it isn’t a swap.

Depth is a separate question

Two tins with the same area and different depths take the same amount of batter to fill to the same level, and they bake completely differently.

Which is why depth doesn’t get folded into the multiplier. The multiplier answers “how much batter”. Depth answers “how will it bake”, and the two need different responses:

Deeper batter insulates its own centre. The outside sets and browns while the middle is still liquid. The usual fix is a longer bake at a slightly lower temperature, around 10–15 °C down, so the heat has time to reach the middle.
Shallower batter bakes faster and dries out sooner. Same temperature, less time, and start checking early.

Keep the same tin and just pour in more batter and you haven’t scaled the recipe. You’ve changed the bake.

The rule of thumb from the baking trade is to fill a tin between half and two thirds. Below half the cake is thin and dries out. Above two thirds it either overflows, or domes and cracks as the outside sets before the middle has finished rising.

Time doesn’t scale with the multiplier

The most persistent misconception here. Doubling a recipe doesn’t double the baking time, and usually barely changes it if the depth stays similar.

Two principles cover nearly everything:

Same depth, larger area means roughly the same time. A tray of cookies twice as wide bakes in the same time as a small one, because heat travels into the batter from the surfaces and that distance hasn’t changed.
Greater depth means considerably longer, disproportionately so. Heat has further to travel, and batter is a poor conductor.

So check for doneness instead of trusting the clock. A skewer, the spring-back test, or an internal temperature (most cakes are done between about 93 °C and 99 °C in the centre) all beat the original recipe’s timing once you’ve changed the tin.

What the multiplier doesn’t handle

Chemical leavening at large factors. Bicarbonate and baking powder don’t scale linearly once you’re far from the original quantity. Too much gives you a soapy or metallic taste and a crumb that rises fast then collapses. Past roughly 1.5×, plenty of bakers hold leavening slightly under the arithmetic answer.

Eggs. They come in whole units, and 1.32 eggs isn’t a quantity. Weigh them, since a large egg is roughly 50 g out of the shell, or nudge the multiplier until it lands on whole eggs and accept a slightly different yield. For home baking the second is usually the better trade.

Very large changes. Past about 2×, splitting across two tins of the original size often beats one much larger tin. You keep the depth, the timing and the leavening behaviour you already know work.

Tins that aren’t the shape you entered. Loaf tins are the worst offender: their sides slope, by different amounts between manufacturers, so the rectangular calculation overstates capacity. Two tins sold at the same nominal size can differ noticeably. If a loaf recipe matters, measure yours by filling it with water and reading the volume.

Why capacity is given in millilitres and never in grams

A tin’s capacity is a volume. Converting that to a weight of batter requires the batter’s density, and densities vary enormously. A whipped sponge and a dense fruit cake of the same volume differ hugely in mass.

Any tool converting a tin size straight to grams of batter is inventing a number, because it can’t know what you’re baking. Volume comparisons are legitimate. Volume-to-weight without an ingredient-specific density isn’t.

Measuring a tin’s real capacity takes a minute and settles the question for good: fill it to the brim with water, then pour that into a measuring jug. That number beats every published figure for your particular tin.

A procedure

Work out both areas. Round is πr² from the diameter; square and rectangular are length × width.
Divide new by old. That’s your multiplier.
Multiply the ingredients. A recipe scaler will do it across a whole list and handle the fractions.
Hold leavening at or slightly below the factor if the change is large.
Decide what you’re doing about the eggs.
Check the depth. Deeper means lower and longer; shallower means shorter.
Fill between half and two thirds.
Ignore the original time and test for doneness.

The arithmetic is the easy part, and it’s the part the converter here does. The judgement, meaning leavening, eggs, depth, and when to use two tins instead of one, is the part no calculator can do for you. That’s why the tool keeps them separate instead of folding everything into one confident number.

Sources and further reading

Last reviewed 2 August 2026.