How Accurate Are Calorie Numbers? Restaurant Labels Can Be Off by 20% — and Your Own Estimate Is Worse

If you count calories to the milligram and wonder why the result doesn't match — this article explains the gap and teaches you how to handle it.

Accuracy of restaurant labels

A 2024 study assessed the accuracy of calorie labelling in England's out-of-home food sector across 295 menu items, finding that overestimation (in 23% of items) and underestimation (in 11%) by more than 20% of measured content were both common, with a mean absolute difference of 21% between reported and measured values^[1].

And worse: your own estimate

Research shows that about two-thirds of people underestimate their meals' calorie content, and a quarter by 500 calories or more^[2]. Human error is far larger than label error.

Even packaged foods: one study found snack energy content was on average higher than labelled, though most items (96%) remained within a 20% margin^[1].

So is tracking pointless?

Quite the opposite. Systematic reviews confirm self-monitoring genuinely supports weight loss (the evidence). The reason: tracking doesn't require absolute accuracy — it requires consistency.

How to handle the error margin intelligently

1. The trend matters more than the number: if your error is consistently 15% every day, week-to-week comparisons remain valid.
2. Calibrate against reality, not the equation: if your weight doesn't move on a logged 2,000 kcal, your actual figure is higher — adjust based on the outcome (how to calculate).
3. Log what you forget: the oil in the pan, sauces and sips of drinks — a bigger error source than label accuracy.
4. At restaurants, estimate high: served portions are usually larger than reference amounts.
5. Use photography: estimating volume against a plate reference reduces the biggest human error — guessing quantity.