Baking Basics

Scoop a cup of flour by dipping the cup into the bag and levelling it, then weigh it. Do it again after fluffing the flour and spooning it in. The two will differ by a substantial margin — commonly 20 percent or more between a packed scoop and a light spoon-and-level — and that gap, repeated across a recipe, is the reason home baking works one week and fails the next with no other change.

Updated 2026-08-28Source: Cereal chemistry references on flour density and hydration, leavening manufacturer technical data, appliance thermostat tolerance specifications
The short versionBuildFigure
Flour, all-purposeAbout 120–125 g per cup, spooned and levelled
Scoop vs spoonSame cup can differ 20%+ by weight
Baking sodaNeeds an acid in the batter to work
Baking powderBrings its own acid. Not interchangeable 1:1.
Oven drift25 °F off is common, more is not rare
The fixA kitchen scale and an oven thermometer

Mass beats volume, and it is not close

Baking is a set of ratios between flour, liquid, fat, sugar, egg and leavening. Change any one of them by a fifth and the result changes — cookies that spread into a sheet or refuse to spread at all, a cake that domes and cracks, muffins that are dry in a way no amount of extra butter fixes. That is the mechanism. Volume measurement is the reason it happens by accident.

Flour is a powder that compresses. A cup dipped into a bag and pressed against the side comes out packed; the same cup filled by fluffing the flour, spooning it in loosely and sweeping the top level comes out light. Depending on technique, humidity and how long the bag has been sitting, the difference lands somewhere between 15 and 30 percent, and a recipe calling for three cups can therefore vary by most of an extra cup between two competent people following the same instructions. Nothing else in home baking produces that much unexplained variation.

IngredientApproximate weight per US cupNote
All-purpose flour, spooned and levelled120–125 gPublished figures vary between sources by a few grams. Pick one and stay with it.
All-purpose flour, scooped and packed145–160 gThe same cup, different technique. This is the whole problem.
Bread flour~127 gHigher protein, more gluten, chewier result
Granulated sugar~200 gDoes not compress much, so volume is more reliable here
Brown sugar, packed~213 gThe recipe means packed, which is why it says so
Butter227 g per cup (2 sticks)One of the few American volume measures that is unambiguous
Water or milk~240 g per cupLiquids are fine by volume
Cocoa powder~85 gCompresses badly. Weigh it.

A cheap digital scale removes all of this and makes cleanup easier as a side effect, since you can tare the bowl and add ingredients into it one after another with no measuring cups involved. Recipes written by weight also scale cleanly: use the recipe scaler to change batch size, the cooking unit converter to move between systems, and baker's percentage to compare two bread formulas that look nothing alike on paper.

What each ingredient is doing

Understanding the job of each component turns a failed batch from a mystery into a diagnosis.

IngredientJobWhat too much or too little does
FlourStructure, from gluten and starchToo much: dry, dense, tough. Too little: spread, collapse, gummy center.
LiquidHydrates the flour, dissolves sugar, becomes steamSteam is a leavening agent in its own right. Wetter doughs rise more and brown differently.
FatTenderizes by coating flour so gluten cannot link up; carries flavorButter temperature matters more than quantity in creamed recipes. Soft-but-cool butter holds air; melted butter does not.
SugarSweetness, browning, moisture retention, tenderizingSugar is hygroscopic and liquefies in the oven, so it drives spread. Cutting it changes texture, not just taste.
EggStructure from protein, richness from yolk fat, some leavening from trapped airWhites tighten, yolks tenderize. An extra yolk gives a richer, chewier cookie; an extra white gives a drier one.
SaltFlavor, and it slows yeastBread without salt tastes flat and ferments too fast. This is not a garnish ingredient.
LeaveningGasSee below. This is where the most common outright errors live.

Baking soda and baking powder do different things

Baking soda is sodium bicarbonate. It is a base, and it releases carbon dioxide only when it meets an acid and moisture. Recipes that call for soda always contain an acid somewhere: buttermilk, yogurt, sour cream, molasses, brown sugar, honey, natural cocoa, lemon juice, vinegar, or fruit. If you substitute a non-acidic ingredient — swapping buttermilk for regular milk, or natural cocoa for Dutch-process, which has been alkalized — the soda has nothing to react with, the rise disappears, and the leftover soda leaves a soapy, metallic taste.

Baking powder is sodium bicarbonate plus a powdered acid plus a starch to keep them apart in the tin. It works in a batter with no acid of its own. Most supermarket baking powder is double-acting, meaning a portion of the gas releases when it gets wet and the rest when it gets hot, which buys you some tolerance for a batter that sits a few minutes before it goes in the oven.

The practical differences: soda is roughly three to four times stronger per teaspoon, it reacts the instant it is wet and in contact with acid, and an excess tastes soapy and metallic while browning the crust more than intended. Powder is weaker per teaspoon, releases part of its gas when wet and the rest in the oven, and an excess tastes bitter and over-inflates before collapsing. They are not a one-for-one substitution in either direction — about 3 teaspoons of powder can stand in for 1 teaspoon of soda, but that also adds salt and acid and changes the result. Check potency before blaming a recipe: a spoonful of soda in vinegar and a spoonful of powder in hot water should both fizz hard.

Both lose potency in an open container. If a tin has been in the cupboard since some previous administration, test it before blaming the recipe.

Your oven is not the temperature on the dial

Home ovens run on a thermostat with a wide tolerance, they cycle above and below the setpoint rather than holding steady, and they drift as they age. Being off by 25 °F is unremarkable and larger errors are common, particularly in older ranges. That is enough to turn a 12-minute cookie into an 8-minute cookie or an 18-minute one, and it explains an enormous share of results that are burnt on the bottom, pale on top, or somehow both.

An oven thermometer that hangs from a rack costs very little and settles the question in twenty minutes. Preheat, let it stabilize for a further 15 minutes after the beep — the beep usually means the air is at temperature while the walls and racks are not — and read it. Then either adjust the dial to compensate or use the calibration offset if the oven has one.

Other things the oven does that recipes assume you know: the back is hotter than the front in most of them, so rotating a sheet halfway through is not superstition. Convection moves air and typically cooks faster, and the usual adjustment is to drop the set temperature by about 25 °F or shorten the time. Dark metal pans absorb more and brown bottoms faster than shiny ones; glass and ceramic heat slowly and hold heat, which is why a glass dish keeps cooking hard after it comes out. Opening the door drops the air temperature quickly, so do it rarely and quickly, and never in the first two thirds of a cake. If you are working from a recipe written in Celsius or need the gas mark, the oven temperature converter handles it.

Failures and their causes

What went wrongMost likely cause
Cookies spread into one flat sheetButter too warm or melted, dough not chilled, too little flour, or a hot pan reused without cooling
Cookies stayed in domes and did not spreadToo much flour — usually a scooped measurement — or too little sugar
Cake domed and cracked down the middleOven too hot; the edges set before the center finished rising
Cake sank in the middleUnderbaked, too much leavening, or the door was opened while the structure was still soft
Muffins and quick breads are tough with tunnels insideOvermixed. Gluten developed. Stir until just combined and accept the lumps.
Everything tastes faintly soapy or metallicExcess baking soda, or soda in a recipe with no acid left in it after a substitution
Bread is dense and did not riseDead yeast, water hot enough to kill it, or salt added directly onto the yeast
Bottoms burn while tops stay paleDark pan, rack too low, or an oven running hot. All three at once is common.
Pie crust shrank down the sidesDough stretched into the pan instead of eased in, and not rested before baking
Results vary week to week with no changesVolume measurement. This is the signature of it.

Habits that fix most of it

Bring butter and eggs to the temperature the recipe assumes, because creaming cold butter never traps air and creaming melted butter has nothing to trap it in. Read the whole recipe before starting, including the part where something needs to chill for two hours. Measure everything out first if the method has any pace to it. Do not double a recipe that involves creaming without checking that the mixer bowl can handle it. Put the leavening in with the flour rather than dumping it into wet ingredients, so it distributes before it starts reacting.

And test for done by something other than the clock: a skewer that comes out with a few moist crumbs rather than wet batter, an internal temperature around 200–210 °F for most breads, edges that have pulled slightly from the pan on a cake. Times in recipes were written for someone else's oven, and now you know how much that is worth.

Questions people ask

How much does a cup of flour weigh?

Spooned into the cup and levelled off, all-purpose flour runs about 120 to 125 grams depending on which source you use. Scooped straight from the bag and pressed, the same cup can hold 145 to 160 grams. That 20 to 30 percent gap is why the same recipe produces different results for different people, and why a scale is the single highest-value purchase in a baking kitchen. Pick one published figure, weigh to it, and your results stop moving around.

Can I use baking powder instead of baking soda?

Not as a straight swap. Baking soda is a base that only produces gas when it meets an acid, so recipes calling for it always contain buttermilk, yogurt, brown sugar, molasses, natural cocoa, vinegar or similar. Baking powder already contains its own acid and works without one. Soda is roughly three to four times stronger by volume, so replacing 1 teaspoon of soda takes about 3 teaspoons of powder, which also adds salt and acid and will change flavor and browning. Going the other direction is worse, because the acid the powder was providing is now missing.

Why do my cookies spread too much?

Usually butter temperature. Butter that was soft enough to look melted, or a dough that went onto the sheet without chilling, gives the fat nothing to hold shape with and the cookie relaxes into a puddle before the structure sets. Reusing a hot sheet pan does the same thing. The other common cause is too little flour, which points back at volume measurement. Chill the dough for at least half an hour, use a cool pan, and weigh the flour before assuming the recipe is wrong.

Is an oven thermometer really necessary?

It is one of the cheapest ways to explain a persistent baking problem. Home oven thermostats commonly run 25 °F off the dial and older ranges can be much further out, and they cycle above and below the setpoint rather than holding a line. That is easily enough to burn the bottoms of cookies or leave a cake raw in the middle at the time the recipe stated. Hang one from the rack, preheat, wait an extra 15 minutes past the beep so the walls catch up, and read what the oven is actually doing.

Why are my muffins tough and full of holes?

Overmixing. Stirring a wet batter develops gluten, and gluten is exactly what you want in bread and exactly what you do not want in a muffin or a quick bread. The tunnels are gas escaping through a structure that has become elastic instead of tender. Combine wet and dry until the flour has just disappeared, leave the lumps alone, and stop. A batter that looks slightly under-stirred bakes into a better muffin than a smooth one.

Related