Are Calorie Counts on Food Labels Accurate? The 20% Rule

This article is for educational purposes only and is not medical advice. Talk with a qualified healthcare professional before making major nutrition, weight loss, or medication-related changes.

Calorie counts on U.S. food labels are close but not exact: federal rules treat a packaged food as misbranded when lab tests find more than 20% more calories than the label states, and in one test of 24 snack foods the median product had 4.3% more. The rule runs one way. A label may state more calories than the food delivers, and feeding studies found people absorb fewer calories from whole almonds, walnuts, and cashews than the standard calculation predicts.

This guide is for anyone who logs packaged or restaurant food and wants to know how far to trust the numbers. It explains the U.S. labeling rule, what lab tests of packaged foods, nuts, and restaurant meals found, and how label error compares with portion error in one day's log. It covers U.S. rules, plus one menu study from England. It does not set a calorie target.

What Does the 20% Rule on Food Labels Actually Say?

Federal labeling rules treat a packaged food as misbranded when a lab sample contains more than 20% more calories than the label declares. The same test in 21 CFR 101.9(g)(5) applies to total sugars, total fat, saturated fat, trans fat, cholesterol, sodium, and, when all of a food's sugar is added, added sugars. A common misreading is that labels may be 20% off in either direction. The text sets a ceiling on excess only.

Four details in the regulation change how to read a label:

• A shortfall is not a violation of the 20% test. Under 21 CFR 101.9(g)(6), "reasonable deficiencies" of calories under the labeled amount are acceptable within current good manufacturing practice. A label can state more calories than the food contains.
• Compliance is judged on a composite of 12 packages. Under 21 CFR 101.9(g)(2), the sample is 12 consumer units, one from each of 12 randomly chosen shipping cases, analyzed together. A single package can sit further from the label than the average of twelve.
• Measurement variation is allowed for. Paragraph (g)(5) also says no regulatory action is based on a value that exceeds the 20% line by less than the recognized variability of the lab method.
• The label describes the gram weight. Under 21 CFR 101.9(g)(7), compliance is based on the metric amount in the serving size statement, not on the cup or piece count.

Protein runs the other way. Under 21 CFR 101.9(g)(4), naturally occurring protein, total carbohydrate, and fiber must measure at least 80% of the declared amount, and nutrients added to fortified foods must measure at least 100%. So the regulation protects you from a label that overstates protein, while the calorie test protects you from a label that understates calories.

The calorie number itself is usually calculated, not measured on your package. 21 CFR 101.9(c)(1)(i) lets manufacturers use specific Atwater factors, the general factors of 4, 4, and 9 calories per gram of protein, carbohydrate, and fat, approved food-specific factors, or bomb calorimetry with a protein correction. Under (g)(8), compliance can also be shown with an approved nutrient database. Rounding adds its own gap, which our guide to why calories don't match the macros on a label works through.

How Close Are Packaged-Food Labels in Lab Tests?

In the U.S. lab tests reviewed here, packaged foods measured a little above their labels on average, well inside the 20% line. The largest test measured 24 popular snack foods with bomb calorimetry in 2010.

• Snack foods. In a 2013 study of snack food label accuracy, the median product had 4.3% more metabolizable calories than its label after adjusting for serving weight, about 6.8 calories per serving. The middle half of products ran from 0.2% to 13.7% over. Servings weighed a median 1.2% more than stated. In 10 products analyzed for macronutrients, carbohydrate ran 7.7% over the label, while fat and protein did not differ significantly. The authors concluded the excess stayed within the federal tolerance.
• Frozen meals. In a 2010 study of reduced-calorie prepared foods, 10 supermarket frozen meals averaged 8% more calories than stated. The difference was not statistically significant because results varied widely between meals.

Both studies are small, and both are more than a decade old. Recipes change, so treat them as a picture of how labels behaved in those samples, not a current rating of any brand. We did not find a newer U.S. lab test of packaged-food labels.

Why Whole Nuts Can Have Fewer Calories Than the Label

Whole nuts are the clearest case where the standard calculation overstates the calories your body takes in. The Atwater factors assume typical digestibility, but in controlled feeding studies at the U.S. Department of Agriculture, people excreted more of the energy in whole nuts than the factors predict. The studies measured metabolizable energy, the calories left after losses in stool and urine.

The studies are the 2012 study of measured energy values of almonds, the 2016 study of available energy from walnuts, the 2018 cashew study, and the 2012 study of the measured energy value of pistachios. The 2012 almond study was co-funded by the Almond Board of California and the cashew study by nut-industry councils, alongside USDA.

The form of the nut matters. In a 2016 study of how food processing and structure affect the metabolizable energy of almonds, whole raw almonds delivered 4.42 calories per gram (about 124 per 28 g), whole roasted almonds 4.86, and chopped almonds 5.04. Almond butter delivered 6.53 calories per gram (about 183 per 28 g), similar to the Atwater prediction. The authors suggested that whole almonds fracture into fewer, larger particles, which may hold back some of their fat.

Are Restaurant Menu Calories Accurate?

Restaurant calories were accurate on average in a large U.S. test, but individual items varied far more than packaged foods did. U.S. menu labeling covers chains with 20 or more locations under the same name, under 21 CFR 101.11(a). Under 101.11(c), a chain needs a "reasonable basis" for its numbers, such as nutrient databases, cookbooks, lab analyses, or Nutrition Facts labels, and must take reasonable steps so preparation and portions match that basis. The section does not include the 12-package, 20% test used for packaged foods.

• United States, 269 items. In a 2011 study of restaurant food energy contents, researchers bought foods from 42 quick-serve and sit-down restaurants in Massachusetts, Arkansas, and Indiana in 2010. The average gap from stated calories was 10 calories per portion, which was not statistically significant. Yet 50 items, 19%, had at least 100 calories more than stated. Foods with lower stated calories tended to contain more than stated, and higher-calorie foods less.
• Reduced-calorie restaurant meals. In the 2010 study above, 29 reduced-calorie restaurant foods averaged 18% more than stated, a difference that was also not statistically significant. Free side dishes raised the calories served to an average of 245% of the stated value for the entree they came with.
• England, 295 items. In a 2025 study of menu calorie labels in England, items sampled in 2024 averaged 16.7 calories below their menu values. Still, 35% of items differed from the measured value by more than 20%: 23% of items listed too many calories and 11% too few. Pubs showed more discrepancies than other outlet types.

For logging, the posted number is still the place to start for a chain meal. Log the exact item and size, then add sides, sauces, and drinks as separate entries. Our guide to tracking calories when eating out covers local restaurants without posted numbers.

Worked Example: How Much Does Label Error Move One Day's Log?

In this example, label error moves the day by 52 to 240 calories, and two portion slips move it by 170. These are fictional labels, not real products. The middle column applies the 4.3% median from the snack-food test to every item. The right column puts every item at the 20% line, a worst case far above that tested median.

Against a planned 500-calorie daily deficit, the 52-calorie gap is about a tenth, and the 240-calorie worst case is about half. The same arithmetic runs in the other direction for foods that deliver less than their label, such as whole nuts.

How to Log Labeled Foods When the Numbers Are Not Exact

Use the label as your reference, and spend your effort on the parts of the log you control.

1. Log the label value as printed. Adding 20% to everything is a guess, because tested snack foods ran a median 4.3% over and whole nuts ran under.
2. Match the label's gram weight. Compliance is judged on the metric serving size, so a food scale puts your portion on the same basis as the label. The food serving converter scales label values to the grams you ate. Without a scale, see our guide to tracking calories without weighing food.
3. Count the servings you actually ate. Check servings per container before logging a whole package as one serving.
4. Keep one entry per food. Switching between different entries for the same product makes week-to-week totals harder to compare.
5. At restaurants, log the posted item and add the extras. Sides, dressings, sauces, and drinks go in as separate entries.
6. Judge your target by your weight trend over several weeks. Daily weigh-ins move with water and food in the gut, as our guide to day-to-day weight fluctuation explains. If the trend and the log disagree, check portions and missing items first, because in the example above they moved the log more than label tolerance did.

Who Should Be Careful With This?

If a clinician or registered dietitian has set a precise intake for a medical condition, ask them how to handle label variation instead of adjusting values on your own. If chasing precise numbers makes eating more stressful, or you have a history of disordered eating, this level of detail may not help you. A registered dietitian can suggest a looser method that still works for your goals.

Where Kalo Fits

Kalo is a calorie and macro tracking app for iOS and Android. For a packaged food, scan the barcode or describe it by text, then review the serving size, portion, calories, and macros before saving, so the entry matches the grams you ate rather than a default serving.

Kalo's numbers are estimates you can edit, so checking them against the package is part of logging. Your daily targets then show what is left today. Kalo is free to download, and using it requires a subscription; current terms are shown on the store purchase screen.

Frequently Asked Questions

Sources

Sources checked October 19, 2026. The 200-calorie table, the six-food day, and the portion examples use fictional labels; the 4.3% and 20% columns are arithmetic, not predictions for a real product. Study results describe the foods those studies tested. Kalo publishes this guide and makes a calorie tracking app. No clinical review is claimed.

• 21 CFR 101.9(c)(1) and (g)(2) to (g)(8): calorie calculation methods, rounding, and nutrition label compliance
• 21 CFR 101.11(a) and (c): covered restaurant chains and the reasonable basis for menu nutrient values
• Jumpertz R, et al. Food label accuracy of common snack foods. Obesity (2013)
• Urban LE, et al. The accuracy of stated energy contents of reduced-energy, commercially prepared foods. J Am Diet Assoc (2010)
• Urban LE, et al. Accuracy of stated energy contents of restaurant foods. JAMA (2011)
• Finlay A, et al. Accuracy of menu calorie labelling in the England out-of-home food sector during 2024: assessment of a national food policy. Br J Nutr (2025)
• Novotny JA, et al. Discrepancy between the Atwater factor predicted and empirically measured energy values of almonds in human diets. Am J Clin Nutr (2012)
• Baer DJ, et al. Walnuts Consumed by Healthy Adults Provide Less Available Energy than Predicted by the Atwater Factors. J Nutr (2016)
• Baer DJ, Novotny JA. Metabolizable Energy from Cashew Nuts is Less than that Predicted by Atwater Factors. Nutrients (2018)
• Baer DJ, et al. Measured energy value of pistachios in the human diet. Br J Nutr (2012)
• Gebauer SK, et al. Food processing and structure impact the metabolizable energy of almonds. Food Funct (2016)