Are Restaurant Calorie Counts Accurate?
Restaurant calorie counts are accurate on average across a whole menu, but not reliably so for any single dish you order. A major US study that tested this directly using lab equipment and found that one in five individual items had at least 100 calories more than stated, and a smaller study found meals specifically marketed as “reduced-energy” or diet-friendly ran higher than stated too, though that particular result didn’t reach statistical significance.
In brief
- A 2011 study lab-tested 269 real restaurant meals and found the average stated calorie count across all of them was accurate, but 19% of individual items had at least 100 calories more than listed.
- That same study found that dishes with a lower stated calorie count were more likely to be understated than dishes with a higher one.
- In a much smaller study (29 restaurant items, 10 frozen supermarket meals) the main finding didn't reach statistical significance because of how much it varied dish to dish. Meals marketed as "reduced-energy" ran 18% higher than stated on average, and when a free side dish came with the entrée, the two together averaged 245% of the entrée's own stated calorie count.
- A 2025 peer-reviewed UK study of nearly 300 menu items under the country's mandatory calorie-labelling law found over a third fell outside the legally permitted margin, in both directions, with pubs the least accurate outlet type overall.
What does the research actually say?
One of the most direct tests of this came from a 2011 study published in JAMA, before the US federal law requiring calorie labels at large chain restaurants took effect in 2018, so it’s worth keeping in mind that labeling practices may have improved somewhat since. Researchers bought 269 food items from 42 restaurants across three US states and measured their actual energy content in a lab (using bomb calorimetry) and compared it against what the restaurant had stated. Averaged across everything, the stated calories were essentially accurate, off by only about 10 calories per portion, a difference the researchers themselves called not statistically meaningful.
That average hides a lot of individual variation, though. Nineteen percent of the 269 items, about 1 in 5, contained at least 100 calories more than what was listed, and the worst-offending tenth of items averaged around 260-290 calories more than stated. So the honest picture isn’t “restaurant menus lie” or “restaurant menus are fine,” it’s that the average across a whole menu can be accurate while any specific dish you actually order has a real chance of being meaningfully off.
Are lower-calorie or “diet” options actually worse?
There’s a genuinely useful, slightly counterintuitive finding buried in that same 2011 study. Dishes with a lower stated calorie count were more likely to have their calories understated than dishes with a higher one. In other words, the “light” or lower-calorie-sounding option on a menu isn’t automatically the safer bet for accuracy.
A separate, earlier, and much smaller study points the same direction, though it’s worth noting it’s now over 15 years old too (e.g., before the 2018 law). In 2010, researchers tested just 29 restaurant items and 10 supermarket frozen meals, all specifically marketed as “reduced-energy.” Restaurant reduced-energy meals measured 18% higher in calories than stated on average, and frozen supermarket meals measured 8% higher, though with a sample this size and this much dish-to-dish variation, the researchers themselves said the result didn’t reach statistical significance. Two more specific findings from that same small study are still worth knowing anyway, some individual restaurant items ran as high as 200% of their stated calories, and when a free side dish came with the entrée, the two together averaged 245% of the entrée’s own stated calorie count. Small and inconclusive statistically, but a real enough pattern that the researchers flagged it as worth taking seriously for anyone managing intake by these numbers.
Is there a legal limit for how wrong a menu can be?
In the UK, where calorie labelling has been legally required in larger food businesses since 2022, the rules allow a 20% margin of error between a menu’s stated calories and the true figure. A peer-reviewed 2025 study tested nearly 300 real menu items against that standard using the same lab-testing approach as the US research, and found just over a third of items, both under and over their stated amount, fell outside that legally permitted 20% margin. Pubs were the least accurate outlet type, and unlike the “diet items get understated” pattern in the older US research, pubs’ calories tended to be overstated rather than understated, a reminder that which direction the error runs seems to depend on the kind of food and outlet, and not a single universal pattern. One caveat on this study’s method is that most items sampled per outlet were picked at random, but each outlet also had a staff-nominated “most popular” item included, so it isn’t a true random sample of everything on every menu.
Does this apply if you’re trying to eat more, not less?
This isn’t a one-directional problem. The research above shows both overstatement and understatement happening, sometimes in the same study. If you’re trying to eat more, an understated “big” meal might actually be helping you eat more than you think, which is exactly what you want if you’re trying to gain, while an overstated one could leave you eating less than you think. If you’re trying to eat less, an understated “light” option, the exact one you might reach for on purpose, is the one this research suggests deserves the least blind trust.
What this actually means for you
None of this is restaurants trying to deceive anyone. The researchers behind these studies point to portion size variation, ingredient swaps, and the practical difficulty of lab-testing every dish exactly as it’s actually served, rather than deliberate misrepresentation. But it does mean that treating a single restaurant meal’s stated calorie count as an exact number, especially the “healthy-sounding” option, isn’t something the evidence actually supports.
This is part of why Ori doesn’t ask you to treat any single logged meal as a precise figure to hit. Your weekly target adjusts based on what’s actually happening with your weight and food trends over time, which naturally absorbs the kind of everyday inaccuracy this research shows is common, rather than expecting one restaurant receipt to be exactly right.
Cooking at home tightens some of this up, though less than you’d think. We went through what a kitchen scale does and doesn’t fix separately.
Sources
- Accuracy of Stated Energy Contents of Restaurant Foods — JAMA (Urban, McCrory, Dallal, Das, Saltzman, Weber & Roberts, 2011)
- The Accuracy of Stated Energy Contents of Reduced-Energy, Commercially Prepared Foods — Journal of the American Dietetic Association (Urban, Dallal, Robinson, Ausman, Saltzman & Roberts, 2010)
- Accuracy of Menu Calorie Labelling in the England Out-of-Home Food Sector During 2024: Assessment of a National Food Policy — British Journal of Nutrition (Finlay, Jones, Thorp, Putra, Polden, Adams, Brealey & Robinson, 2025)
A gentle safety note
Ori supports general wellness. It is not medical care. If you have a history of disordered eating, are pregnant or breastfeeding, take medicines that affect weight or appetite, or have a clinical nutrition need, speak with a qualified healthcare professional before changing your intake or weight goal.
Read Ori's Health Disclaimer