According to what I've seen, fat has a listed caloric value of 9 Cal/g, while carbs and protein have listed caloric values of 4 Cal/g.* Are these numbers exact, or have they been rounded? And if they have been rounded what are the exact numbers? One hit said that the numbers are rounded, but not by how much.

*These are "food calories" (aka "large calories" or "Calories" with a capital "C").
1 food calorie = 1000 "small calories" or "gram calories" = 1 kcal = 4.184 kJ

See: https://en.wikipedia.org/wiki/Calorie

  • $\begingroup$ I have burnt in my memory from deep past rather 4.3 kcal/g for "carbs" and 9.1 kcal/g for fats. So yourvalues are definitely rouded up, covering the value span across particular food components. $\endgroup$
    – Poutnik
    Aug 31, 2020 at 7:08

2 Answers 2


They're not exact numbers. These numbers aren't exact for three reasons:

  1. Each type of carb, protein, and fat has a different caloric value. These are overall averages for each class.

  2. Even if you were dealing with a single pure compound, the value couldn't be exact because there is individual variation in how much of that compound is metabolized based on digestion, absorption, etc.

  3. Even if #1 and #2 were not issues, the number still couldn't be exact, because it is a measured value, and measured values are never exact. Only counted numbers, and values that are exact by defintion (e.g., the speed of light in SI units) are exact.

To get a better idea of how these values are determined, and the issues and uncertainties associated with measuring them, I'd recommend reading through:

Food and Agriculture Organization of the United Nations. FAO food and nutrition paper. Chapter 3: Calculation of the energy content of foods - energy conversion factors. Food and Agriculture Organization of the United Nations, 1979.

Here is a direct link to chapter 3:


  • 4
    $\begingroup$ Not only that, but a saturated triglyceride containing three 16:0 fatty acid chains will have a (slightly) different caloric value per gram than one containing, say, two 16:0 chains and one 14:0 chain. And, likewise, different carbs will have different caloric values. Whether such differences, particularly if they are small, are nutritionally significant, is an entirely different question. But you weren't asking about nutritionally significant differences, you were asking if the numbers are exact. $\endgroup$
    – theorist
    Aug 29, 2020 at 3:28
  • 8
    $\begingroup$ @ToddMinehardt Nope, I'm not off by a factor of 1000. First, it's understood from context that these are nutritional Calories (=1000 calories). The OP's mistake was just in not capitalizing the "C". Second, I never specified either "calorie" or "Calorie" in my answer. Third, and most importantly, anyone who carefully thinks this through will see that your "gotcha" is what's immaterial since, logically, his questions—about whether these values are exact, and whether there are within-class differences between these values—are independent of how the unit itself is scaled. $\endgroup$
    – theorist
    Aug 29, 2020 at 4:38
  • 1
    $\begingroup$ @Constantthin The effects of low-fat vs low-carb diets is a very complicated and unresolved issue in the field of nutrition. I took a quick look through the primary source (cell.com/cell-metabolism/fulltext/S1550-4131(15)00350-2) cited in the popular press article you linked and, IME, to the extent they are seeing differences it is not because they used an incorrect figure for the caloric value of fat. If you read the primary source you can see there are many metabolic differences created by the different diets, and that's much more likely to be the explanation. $\endgroup$
    – theorist
    Aug 29, 2020 at 6:22
  • 1
    $\begingroup$ @Constantthin: What's the point of counting calories if you're just going to make arbitrary adjustments? That's not counting, that's guessing badly on purpose. $\endgroup$ Aug 29, 2020 at 17:36
  • 2
    $\begingroup$ ... Ask yourself this: Is there any way—any way at all—in which my answer would need to be different if the questioner expressed those values as cal instead Cal? [I.e., if he instead referenced 9000 cal/g for fats, and 4000 cal/g for carbs and proteins.] The answer is clearly no, demonstrating the compete irrelevance of the scaling factor to the question at hand. (5) Thanks for the positive comment on my answer :). $\endgroup$
    – theorist
    Aug 31, 2020 at 21:45

It doesn't "take one calorie to burn", you gain one calorie by burning.

Now noting obviously what everybody else has said about selection of the type of fat affecting the precise amount of energy available from one gram (which in practice will be measured in a calorimeter), but what this is really saying is that by metabolising some specified fraction of a gram of fuel you can do (approximately) one calorie of work, e.g. raising the temperature of water by some specified amount or moving your muscles in such a way that they lift a weight by some specified number of centimeters.

So no, it's not an integer: it's an approximation to the amount of work you'd have to do to prevent the food you've just eaten being stored as flab, and to get an accurate figure you'd have to do an assay of an identical portion of that food taking into account both its chemical composition and its digestibility (e.g. an increased amount of fibre in the specimen might not change the result of the physical calorimetric test, but might change the availability of the nutrients to the digestive system).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.