Calorie counts on food labels come from a system developed in the late 1800s that assigns a fixed energy value to each gram of protein, fat, and carbohydrate in a food. Rather than burning every granola bar in a laboratory, manufacturers use a simple math formula: multiply the grams of each macronutrient by its standard calorie factor, then add them up. The system is practical and widely used, but it’s also an approximation, and the real number of calories your body extracts from a food can differ from what’s printed on the package.
The Original Method: Burning Food in a Lab
The foundational technique for measuring food energy is called bomb calorimetry. A dried food sample is placed inside a sealed steel chamber, pressurized with pure oxygen, and ignited. The chamber sits inside a jacket of water, and scientists measure exactly how much the water temperature rises during combustion. That temperature change reveals the total chemical energy stored in the food, measured in calories.
This method captures gross energy, meaning every last bit of chemical energy the food contains. But your body isn’t a furnace. You don’t combust food completely. Some energy is lost during digestion, some passes through you unabsorbed, and protein burns differently in living tissue than it does in a steel chamber. So while bomb calorimetry gave scientists a starting point, it took additional work to figure out how much energy humans actually use from what they eat.
The Atwater System: 4, 9, and 4
In the 1890s, a chemist named Wilbur Atwater tackled exactly that problem. He combined bomb calorimetry data with studies of human digestion, measuring what went into people’s bodies and what came out, to estimate how much energy each macronutrient actually delivers after accounting for digestion losses. The result was a set of average conversion factors still used today: protein provides 4 calories per gram, carbohydrates provide 4 calories per gram, and fat provides 9 calories per gram. Alcohol, when present, is counted at 7 calories per gram.
These numbers, known as the Atwater factors, are averages across many different foods. In 1955, researchers refined the system by calculating specific factors for individual food groups, recognizing that not all proteins, fats, and carbohydrates are created equal. The protein in rice, for example, has a heat of combustion roughly 20 percent higher than the protein in potatoes. Whole wheat flour and finely milled white flour deliver different amounts of available energy even at the same weight, because milling removes fiber and makes starch easier to digest. Still, the general 4-9-4 rule remains the default for most food labeling worldwide.
How Calorie Counts End Up on Labels
Most calorie counts you see on packaging aren’t measured by burning that specific product. Instead, manufacturers analyze the food’s composition, determining how many grams of protein, fat, carbohydrate, water, and ash (mineral content) it contains. Carbohydrate is usually calculated “by difference,” meaning scientists measure everything else and whatever percentage of the food’s weight remains is classified as carbohydrate.
Once those macronutrient amounts are known, the Atwater factors do the rest. If a serving contains 10 grams of protein, 5 grams of fat, and 30 grams of carbohydrate, the math is: (10 × 4) + (5 × 9) + (30 × 4) = 40 + 45 + 120 = 205 calories. That’s the number that goes on the label.
The USDA’s FoodData Central database, which provides the nutrient data behind thousands of products and ingredients, works the same way. Trained labs perform chemical analyses on representative samples of foods, and energy values are calculated from the protein and fat results using the general Atwater factors. This database feeds into nutrition apps, restaurant calorie counts, and dietary guidelines.
How Accurate Are Label Calories?
The FDA allows a food’s actual calorie content to exceed its label claim by up to 20 percent. That means a snack bar labeled at 200 calories could legally contain up to 240 calories. Research testing common snack foods has found that most products fall within this 20 percent window, but the tolerance is notably one-directional: the FDA is primarily concerned about products containing more calories than stated, not fewer.
Beyond labeling tolerance, the Atwater system itself introduces imprecision. It tends to overestimate how many calories your body actually absorbs, particularly in foods high in fiber or protein. The system assumes a relatively consistent level of digestibility across foods, but real digestion is messier than that.
Why Your Body Gets Fewer (or More) Calories Than the Label Says
The calories printed on a package represent a theoretical maximum of metabolizable energy. In practice, the physical structure of a food changes how much energy you actually absorb. Whole almonds are a striking example. Studies tracking body weight in people who added almonds to their diets found that participants gained far less weight than the added almond calories would predict. In one study, subjects consumed roughly 57,500 extra calories from almonds over the study period, which should have produced about 6.4 kilograms of fat gain. The actual average gain was just 0.4 kilograms. Part of this was because people naturally ate less of other foods to compensate, but part of it was that the intact cell walls of whole almonds physically block some fat from being digested and absorbed.
Cooking pushes things in the opposite direction. Heating food, whether it’s meat or sweet potato, increases caloric availability by breaking down fiber and denaturing proteins so your gut can extract more energy. Processed peanut butter releases more absorbable calories than the same weight of whole peanuts. Finely milled flour is easier to digest than coarsely ground grain. Foods with similar macronutrient profiles on paper can deliver meaningfully different caloric payoffs depending on how they’ve been prepared.
This means the same 170-calorie serving of almonds delivers different amounts of usable energy depending on whether you eat them whole, sliced, or ground into butter. The label can’t capture that variability.
Special Cases: Fiber, Sugar Alcohols, and Organic Acids
Not everything classified as “carbohydrate” delivers the full 4 calories per gram. Dietary fiber passes through your digestive system largely intact, so it contributes far fewer calories than sugar or starch. Sugar alcohols, commonly found in sugar-free candies and protein bars, provide roughly half the calories of regular sugar. Organic acids like citric acid contribute about 3 calories per gram. Some food labels account for these differences by subtracting fiber and sugar alcohol calories from the total, while others don’t, which is one reason calorie counts for similar products can vary between brands.
The Bottom Line on Precision
Calorie counts are useful estimates, not exact measurements. They rely on a 130-year-old system of average conversion factors applied to chemically analyzed macronutrient content. The system works well enough to guide broad dietary decisions, but it doesn’t account for food structure, cooking method, individual variation in gut bacteria, or differences in chewing thoroughness. For most people, label calories are close enough to be genuinely helpful. They’re just not the precise accounting system they appear to be.