Why Your Recipe Conversions Are Off: Density and Ingredient Explained

Inches and centimetres both measure length, so one converts to the other by a fixed number. That is what most people expect a unit conversion to be, and it is why the grams-to-teaspoons question feels like it ought to have a single answer.

It does not, because grams and teaspoons measure different physical quantities. A gram is a quantity of matter. A teaspoon is a quantity of space. Asking how many grams are in a teaspoon is like asking how many kilograms are in a suitcase — it depends entirely on what you packed.

The bridging quantity is density: how much mass a substance packs into a given volume. Multiply density in grams per millilitre by the 4.93 millilitres of a US teaspoon and you get grams per teaspoon. That single multiplication is the whole of the conversion, and the density is the whole of the problem.

Here is the subtlety that makes the numbers counter-intuitive.

When you fill a spoon with a powder, you are not filling it with solid material. You are filling it with particles and the air between them. What matters is bulk density — the mass of the pile divided by the volume the pile occupies — which is always lower than the density of the material itself, sometimes by a lot.

Salt is the cleanest demonstration, because the material never changes. Table salt, Morton kosher and Diamond Crystal kosher are all sodium chloride, chemically identical, with the same true density. Yet a teaspoon holds 6.0 g of table salt, 4.8 g of Morton, and only 2.8 g of Diamond Crystal. Table salt's tiny cubes stack with almost no gaps. Diamond Crystal's hollow pyramids stack loosely and trap air in the hollows. Same substance, 114% more weight in the same spoon, purely from crystal geometry.

The same logic explains why fine powders feel light but are not uniformly so. Cocoa at 1.8 g per teaspoon is light because its irregular particles cling to one another and build an open structure around trapped air. Baking soda at 4.8 g is also a fine powder, but its particles are compact and they settle rather than bridge.

The error is not random noise. It is systematic — it pushes the same direction every time, by the same proportion — which is exactly the kind of error that a cook cannot detect by averaging over several bakes.

Take a recipe needing 10 g of active dry yeast. The correct conversion is 3.23 teaspoons. Run it through a converter that quietly assumed granulated sugar and you get 2.38 teaspoons — you would use 26% too little yeast, every single time, and the dough would underprove every single time.

Or 10 g of cocoa powder. Correct: 5.56 teaspoons. Assuming water instead: 2.04 teaspoons — barely a third of the cocoa the recipe wanted.

Or 10 g of table salt. Correct for table salt: 1.67 teaspoons. Apply that number to Diamond Crystal kosher and you have put in 47% of the salt intended, which reads as bland rather than wrong and is therefore never diagnosed.

Even with the right ingredient selected, a reference figure is an average over conditions that vary. Four things move it:

Liquids are the exception that proves the rule, and it is worth understanding why, because it is the reason so many generic converters get away with having no ingredient selector at all.

A liquid has no packing state. There are no particles and no air gaps, so its bulk density is simply its density, and it does not change with how you pour it. Water is 1 gram per millilitre by definition, which makes a teaspoon 4.9 g and a cup 237 g.

Better still, most kitchen liquids sit close to water. Milk is 5.1 g per teaspoon, barely 4% heavier because of its dissolved sugars and proteins. Cooking oils cluster near 4.5 g — olive, sunflower, rapeseed and groundnut are within a percent or two of each other, so oils are effectively interchangeable by volume. Vanilla extract at 4.3 g is slightly lighter than water because it is alcohol-based.

Honey and syrups break the pattern: honey is 7.0 g per teaspoon, 43% heavier than water, because it is mostly dissolved sugar with very little free water. Golden syrup, molasses and maple syrup all sit in the same heavy range.

So a converter with no ingredient selector is roughly right for water, milk, stock, wine and juice, and wrong for everything else in the kitchen. That is a narrow band of correctness dressed up as a general tool.

There is a pattern to which conversion errors ruin food, and it is not the one people expect.

Flour is measured in cups and is the most technique-sensitive ingredient there is, yet a 5% flour error is invisible. There is so much of it, and the recipe has so much slack, that the batter absorbs it.

Yeast is measured in teaspoons, and a 25% error is the difference between a loaf and a brick. Salt is measured in teaspoons, and a 100% error is the difference between dinner and the bin. Baking soda is measured in quarter teaspoons, and a 50% error tastes of soap.

The ingredients where density assumptions do the most damage are precisely the ones small enough to be measured with a spoon in the first place — which is to say, the ones where you were relying on the conversion rather than on your eye.

Three rules cover nearly all of it.

Never use a converter that does not ask what you are measuring. If there is no ingredient selector, one has been chosen for you and you cannot see which.

Check the answer against a band. Dry baking ingredients almost all fall between 1.8 g and 6.0 g per teaspoon. If a tool tells you a teaspoon of something weighs 12 g, it is wrong or it is honey.

Weigh the things that are ratios. Bread, brines, cures and anything expressed as a percentage should never be converted at all — the recipe already gave you the units that work.