Net Carbs, Explained: What a Nutrition Label Tells You and What It Doesn't

Put the same tub of yogurt into two different low-carb apps and you can get two different net carb figures. Neither app is malfunctioning. They are applying different conventions to a term that no US regulator has ever defined.

That is the part most explanations skip, and it is the part that actually helps. Once you know what is on a label and what is not, the arithmetic takes about four seconds and you stop being surprised by disagreements.

What a US Nutrition Facts label actually declares

Under FDA's labeling rules, the carbohydrate portion of the panel is a short list:

"Net carbs" is not on that list. It is not a nutrient FDA defines for the Nutrition Facts panel, and it is not something a manufacturer is required or permitted to declare as part of it. When a net carb figure appears on the front of a box, it is the brand's own arithmetic printed in its own marketing space — which does not make it wrong, but does mean nobody is checking the method.

The Daily Values used for the percentage column are 275 grams of total carbohydrate and 28 grams of dietary fiber, both based on a 2,000 calorie reference diet. FDA is explicit that this reference diet is a common yardstick for comparing products, not a recommendation matched to your own calorie needs.

Why fiber sits inside total carbohydrate

This trips people up constantly, and there is a concrete regulatory reason for it.

On a US label, total carbohydrate is generally not measured directly. It is calculated by difference: the lab measures protein, total fat, moisture and ash, and whatever weight is left over is called carbohydrate. Fiber is part of that leftover weight, so it is structurally inside the total carbohydrate figure rather than sitting next to it.

That single fact explains the entire practice. Fiber is counted in the total, most of it is not digested and absorbed the way starch and sugar are, and so people who care about digestible carbohydrate take it back out again. Net carbs is the name for taking it back out.

The subtraction, and the two conventions

The arithmetic itself is trivial:

net carbs = total carbohydrate − dietary fiber
net carbs = total carbohydrate − dietary fiber − sugar alcohols

The first is the conservative version. The second is the convention most commonly used in US low-carb practice, because sugar alcohols are also inside total carbohydrate and are also handled differently by the body from starch and sugar.

As an illustration, take a label reading Total Carbohydrate 23 g, Dietary Fiber 9 g, Sugar Alcohol 6 g. Fiber only gives you 14 g. Fiber plus sugar alcohols gives you 8 g. Same food, same label, same day — a 6 gram gap that came entirely from which line you decided to count.

Neither answer is cheating. What matters is that you know which one you are looking at.

"Fiber" is not one substance

Since the 2016 label update, dietary fiber on a US label has a specific regulatory meaning: naturally occurring non-digestible carbohydrates that are intrinsic and intact in plants, plus isolated or synthetic non-digestible carbohydrates that FDA has determined have a beneficial physiological effect. Isolated fibers that have not cleared that bar do not get to appear on the Dietary Fiber line.

The practical consequence is that the fiber number on a label is already a filtered figure. Two similar-looking products can declare different fiber amounts partly because of what was added, not only because of what is in the food.

Sugar alcohols are not one substance either

FDA's labeling regulations do not treat all sugar alcohols as interchangeable. For calorie calculation they assign a specific value to each one: erythritol is assigned 0 calories per gram, mannitol 1.6, maltitol 2.1, xylitol 2.4 and sorbitol 2.6, among others.

Look at that spread and the working rule reveals itself for what it is. Subtracting every gram of every polyol treats erythritol and sorbitol as the same substance, and the regulations that govern the label say they are not. It is a defensible simplification, not a measurement. If you subtract all sugar alcohols in full and your numbers feel optimistic, this is where the optimism is coming from.

Allulose and monk fruit are not sugar alcohols at all

Worth flagging because low-carb products lean on them heavily. Allulose is a sugar, not a polyol. Monk fruit sweetness comes from mogrosides, a different class of compound again. An app that subtracts erythritol, allulose and monk fruit is applying a broader convention than "total carbs minus fiber minus sugar alcohols" — a defensible one, and a different one.

Why two apps disagree about the same food

Three ordinary reasons, none of them a bug:

1. Different conventions. One subtracts fiber only, another subtracts fiber plus polyols, a third also subtracts newer sweeteners.
2. Different underlying records. A packaged food's own label and a generic reference entry in a database such as USDA's FoodData Central describe similar foods with different numbers, because one is a specific product and the other is a representative average.
3. Different starting points after rounding. This one is underrated, so it gets its own section.

Rounding is quietly doing a lot of work

FDA's regulations round the carbohydrate figures on a label. Amounts are declared to the nearest gram, an amount below 1 gram may be stated as "less than 1 gram", and an amount below half a gram may be declared as zero.

Read that last clause again with a low-carb pantry in mind. A sauce, a stick of gum, a seasoning packet and a splash of cream can each legitimately declare 0 g and still contribute real carbohydrate to your day. Nothing has been misdeclared. The rounding rule is doing exactly what it was designed to do, and the effect only becomes visible when you eat a lot of small servings.

If you are counting to a tight daily figure, this is the single most common source of the gap between what you logged and what you ate.