Every few months, a headline confirms what you already suspect: the calorie counts on food labels are not accurate. A protein bar gets pulled apart by an independent lab. A restaurant meal comes back with a number the menu never promised.
Real measurements back it up. Bomb calorimetry — literally burning the food to measure its energy — put those suspicions to the test. Out of 269 restaurant meals analyzed, nineteen percent contained at least 100 more calories than the menu stated. Lower-calorie dishes, the ones a dieter picks to stay on track, were the most likely to overcount.
How Accurate Are Calorie Counts on Food Labels?
Calorie labels are imprecise but functional — restaurant meals miss by about ten calories on average, though individual items can be off by a hundred or more. The larger error is yours: people consistently underreport their own intake by 11 to 47 percent. Tracking works anyway, because awareness — not precision — is what drives behavior change.
— Urban et al. 2011 · JAMA · n=269; Lichtman et al. 1992 · NEJM · n=10
On average, those 269 meals were off by ten calories — a gap so small it was not meaningful. Some individual items missed by hundreds. But restaurant menus sit in a regulatory gray zone where no federal rule requires accuracy testing for individual dishes. And even the worst label error does not explain what happens when someone tracks every bite, hits their number every day, and still watches the scale refuse to move.
Label error is real. It is not the error that breaks a diet. People who reported eating 1,200 calories a day were actually consuming over 2,200 — confirmed through doubly labeled water, a method that measures metabolic output instead of trusting food diaries. The gap was 47 percent — more than a thousand calories a day, unaccounted for.
Metabolisms in the same group measured normal — within five percent of predicted values for their body size. No metabolic exception hiding the difference. No uniquely efficient digestion making the calories vanish. Just a persistent, invisible gap between what people remember eating and what they actually eat.
BLAMED: The calorie label
ACTUAL: The person reading it
Across 59 studies and more than 6,000 adults, this pattern shows up in every dietary tracking method ever tested — food diaries, recall interviews, apps. Underreporting ranges from 11 to 47 percent, depending on the method and the population. Everyone does it. The only question is by how much.
So if the label is imprecise and the person reading it is worse — does tracking even work? It does. Not because the numbers are right, but because paying attention changes what you eat next. People who track with imperfect tools — labels, apps, rough estimates — still shift their behavior. Precision was never what made it work.
The label is not what needs fixing. Neither is your metabolism. What actually shifts when you start tracking — and why it works even when every number in the system is off — has less to do with math than you would think.