Why does my recipe not work when I convert it?
Volume measures vary by country and by how you scoop. A cup of flour ranges from 120 to 150 grams depending on technique alone, which is why weighing is the only repeatable method for baking.
Updated 2026-08-22
The two ounces
This causes more ruined recipes than any other conversion, and the units are not even related. An ounce measures mass. A fluid ounce measures volume. They share a name and nothing else.
They coincide only for water, and only approximately — a US fluid ounce of water weighs about 1.04 ounces. For anything else the relationship depends entirely on density. Eight fluid ounces of flour is a cup and weighs roughly 125 grams; eight ounces of flour is 227 grams, nearly twice as much. A recipe that means one and is read as the other is not slightly off, it is wrong by a factor approaching two.
American recipes freely use both, often writing each as "oz" and leaving the reader to infer from the ingredient. The working rule is that anything poured is volume and anything scooped, cut or weighed is mass — but the ambiguity is real and occasionally cannot be resolved from the text.
The fluid ounce itself comes in two sizes. The US one is 29.57 millilitres and the imperial one is 28.41, about four percent apart, because they were defined as fractions of two different gallons. Across a full recipe that four percent accumulates, though it is a much smaller problem than confusing volume with mass.
The tablespoon problem
There are three tablespoons in current use and one of them is meaningfully different. The United States uses 14.79 millilitres, most metric countries use exactly 15, and Australia uses 20.
The first two are close enough to be interchangeable without harm. The Australian tablespoon is a third larger, which means following an Australian recipe with a 15 millilitre spoon under-measures by twenty-five percent, and following an American one with a 20 millilitre spoon over-measures by a third.
Whether that matters depends entirely on the ingredient. A third more oil or stock is survivable. A third more baking powder, bicarbonate of soda, salt or yeast is not, because those are the ingredients measured in single spoons and the ones where proportion decides whether the result is edible.
The teaspoon relationship differs too. Three teaspoons make a tablespoon everywhere except Australia, where the 5 millilitre teaspoon means four fit into the larger tablespoon. Worth knowing before halving a recipe by counting teaspoons.
Cups are not a measurement
A cup is a volume, and dry ingredients do not have a fixed density, so a cup of anything powdered is a range rather than a number.
A cup of flour can weigh anywhere from about 120 to 150 grams depending on nothing but technique. Scooping the cup directly into the bag compacts the flour and gives the high end. Spooning it in loosely and levelling gives the low end. Sifting first gives less again. The same person can produce a twenty-five percent spread across a week without noticing, which is why two people following the same recipe get different results and each assumes they did something wrong.
Brown sugar is worse, because recipes specify it packed, which is an instruction to compress it by an unspecified amount. Chopped ingredients are worse still — a cup of chopped walnuts depends on how coarsely they were chopped.
The cup itself is also not one size. The US customary cup is 236.6 millilitres, the US legal cup used on nutrition labels is 240, and the metric cup used in Australia and much of Europe is 250. Three measures, one name, a six percent spread before any scooping variation is added.
What to do about it
Weigh dry ingredients. This is the entire answer to most of the above, and it is why professional and published baking recipes specify grams. A gram is a gram regardless of how it was scooped, which country the book came from, or how coarsely something was chopped. It is also faster, because a scale lets you add ingredients to one bowl without measuring cups to wash.
For liquids, volume is fine. Water, milk, oil and stock have consistent density and pour without compacting, so a measuring jug is as repeatable as a scale. Millilitres avoid the two-fluid-ounce problem entirely.
Where a recipe gives only cups, convert using a table for that specific ingredient rather than a single factor, since flour, sugar, butter and oats all differ. And prefer the source country's convention: an American recipe means US cups and spoons, and reading it with metric ones introduces error in the same direction throughout.
Butter deserves a paragraph of its own, because American recipes measure it in sticks and nowhere else does. A US stick is a quarter-pound — 113 grams, half a cup, or eight tablespoons — and packaging is printed with tablespoon markings so it can be cut rather than measured. Elsewhere butter is sold in 250 gram blocks with no markings at all, which is why a recipe calling for three quarters of a stick is genuinely awkward outside the country it was written in. Weighing sidesteps it entirely.
Eggs are the quiet variable in baking. A recipe assumes a size, and the sizes are not aligned internationally: a US large egg is about 50 grams of egg without shell, a UK large is nearer 60, and Australian sizing differs again. In a cake using four eggs that is forty grams of difference in liquid and protein, which is enough to change the crumb. Recipes that matter specify a weight or a count of a named size for exactly this reason.
One last measure worth knowing: oven temperatures. Fan or convection ovens run hotter than the conventional figure a recipe assumes, and the standard adjustment is to drop the temperature by twenty degrees Celsius or about thirty-five Fahrenheit. Recipes rarely say which they mean, and a cake baked twenty degrees too hot browns before it sets.