Baking Soda Measurements: Teaspoons, Grams, and Why Precision Matters

Baking soda — sodium bicarbonate, sold in Britain and much of the Commonwealth as bicarbonate of soda — has a bulk density of about 0.97 grams per millilitre. A US legal teaspoon is 4.93 millilitres, so a level teaspoon comes to 4.8 g.

That makes it heavy for a white powder. Compare it with the other pale powders in the same cupboard: all-purpose flour is 2.6 g per teaspoon, cornstarch 2.5 g, powdered sugar 2.5 g. Baking soda weighs nearly twice as much as any of them from the same spoon, because its fine crystals pack densely rather than trapping air the way milled starches do.

It is also noticeably heavier than baking powder, which is 4.0 g per teaspoon. That difference is not incidental — baking powder is baking soda cut with an acid salt and a starch filler, so a spoonful of it contains substantially less of the active ingredient in two separate ways: less mass, and a smaller fraction of that mass doing the leavening.

Baking soda is a base. It does nothing on its own; it reacts with an acid in the batter — buttermilk, yogurt, brown sugar, molasses, cocoa, honey, vinegar, citrus — and that reaction releases carbon dioxide, which is what lifts the crumb.

The reaction is fast and it is stoichiometric: a given weight of soda will neutralise a given quantity of acid and no more. Once the acid is used up, any remaining soda simply sits in the batter unreacted. Unreacted sodium bicarbonate tastes soapy and metallic, and it raises the pH of the crumb, which darkens the colour and can leave a slick, unpleasant texture.

This is why the doses look so small next to everything else in a recipe. A typical cake uses about a quarter teaspoon of baking soda per cup of flour — that is 1.2 g of soda against roughly 125 g of flour, a ratio of about one part in a hundred. There is no headroom in a measurement that small.

Put the error in perspective. If you scoop rather than level and end up with a heaped quarter teaspoon holding half again as much, you have added about 0.6 g extra. That sounds trivial. It is a 50% overdose of the single most reactive ingredient in the bowl.

To convert grams of baking soda to teaspoons, divide by 4.8. For baking soda teaspoons to grams, multiply by the same figure. The amounts people look up most often:

This is the substitution people attempt most and get wrong most. The two are related but they are not the same product, and swapping them one-for-one changes both the lift and the flavour.

Baking powder is baking soda plus a dry acid plus starch. It brings its own acid, so it works in batters that contain none. It is also diluted: roughly a quarter of a commercial baking powder by weight is active sodium bicarbonate. That is where the familiar "baking soda is about four times stronger" rule comes from.

Turning that into numbers: to replace 1 teaspoon of baking powder (4.0 g) you need about a quarter teaspoon of baking soda (1.2 g) — plus an acid, because the soda alone has nothing to react with. Half a teaspoon of cream of tartar per quarter teaspoon of soda is the standard fix.

Going the other way is harder. Replacing baking soda with baking powder means using roughly four times as much by volume, which introduces a great deal of extra starch and acid salt into the batter. It works in an emergency for small quantities and it will noticeably change the taste and texture of anything that needed more than half a teaspoon of soda.

Because the dose is set by the acid available rather than by taste, there is a defensible arithmetic behind the quantities, and knowing it lets you spot a typo or repair a recipe that has been scaled badly.

The general working figure is about ¼ teaspoon of baking soda — 1.2 g — per cup of flour, which is roughly 125 g. That puts soda at about 1.0% of the flour weight, and almost every chemically leavened batter lands within a whisker of that.

The acid side has its own conventions. Half a teaspoon of soda — 2.4 g — is about what one cup of buttermilk or plain yogurt will neutralise. A cup of natural cocoa powder, which is acidic, will take roughly the same. Dutch-process cocoa has been alkalised and neutralises almost nothing, which is why recipes using it lean on baking powder instead.

If a recipe calls for more than about half a teaspoon of soda per cup of flour and has no obvious acid — no buttermilk, yogurt, sour cream, molasses, brown sugar, natural cocoa, vinegar or citrus — treat the quantity as suspect. Either the acid is missing from the ingredient list or the figure should have been baking powder at 4.0 g per teaspoon.

The corollary is that you cannot simply increase soda to get more rise. Past the point where the acid runs out, extra soda contributes nothing but soapy flavour and a darker, coarser crumb. More lift comes from adding baking powder, which brings its own acid, or from mechanical aeration.

Assuming you own a quarter-teaspoon spoon and use it, three things still go wrong.

The first is caking. Baking soda absorbs moisture from the air and forms lumps in the box. A spoonful lifted out of a caked box contains solid clumps and voids around them, and it can weigh anything from 15% under to 20% over the reference figure. Stir or sieve the soda in its container before measuring — the same discipline you would apply to flour.

The second is scooping. Pressing the spoon down into the box compacts the powder against the walls of the spoon. Baking soda compacts more readily than flour does because its particles are fine and roughly spherical. Dip a spoon into the box and drag it out heaped, and you can carry 40% more than a levelled spoon.

The third is the spoon itself. Measuring spoon sets are manufactured to loose tolerances, and cheap ones are routinely 5-10% off nominal. That matters less for a teaspoon of vanilla than it does for a quarter teaspoon of soda, where the same percentage error lands on the ingredient with the least tolerance.

For a single batch of cookies, a levelled quarter teaspoon is fine and the 1.2 g figure is a curiosity. Volume measurement is accurate enough because the recipe has slack in it and one batch is one batch.

It stops being fine when you scale. Triple a recipe and a 20% spoon error becomes 0.7 g of surplus soda spread through a batter that has a fixed amount of acid to neutralise it — that surplus has nowhere to go. It also stops being fine for anything you intend to reproduce exactly, or for chemical-leavened breads and soda breads where the soda is the entire leavening system.

A 0.1 g resolution scale — the sort sold for coffee or for jewellery — resolves a quarter teaspoon of baking soda into about a dozen increments. That is the difference between a measurement you can repeat and one you can only approximate.