Should you weigh food raw or cooked?
Weigh raw, and log against a raw entry. Raw rice is listed at 365 kcal per 100 g and cooked rice at 130 kcal per 100 g[1]; raw chicken breast at 120 kcal per 100 g and roasted at 165[2]. The energy did not change, the water did. Logging 200 g of cooked rice against the raw entry over-records one side dish by 470 kcal.
On this page
Cooking moves water, not calories
Heat does not add or remove energy from food in any amount that matters. It moves water — into rice and pasta, out of meat — and water weighs something while contributing nothing. Every mismatch in this article comes from that one fact.
Rice is the clearest case. Raw white long-grain rice is listed at 365 kcal per 100 g; the same rice cooked without salt is listed at 130 kcal per 100 g[1]. The grain did not lose energy. It absorbed water, so each 100 g of what is now on the plate contains less of the original grain.
Meat runs the other way. Raw skinless chicken breast is listed at 120 kcal per 100 g and the same cut roasted at 165 kcal per 100 g[2] — the water left, so what remains is more concentrated. Cooking yields for meat and poultry are measured and published for exactly this reason, because the weight change is large enough to matter in a nutrient database[3].
The mistake this causes, in numbers
Suppose you cook rice, put 200 g on the plate, and log it against the raw entry. At 365 kcal per 100 g you have recorded 730 kcal. What you actually ate was 200 g of cooked rice, which is 260 kcal[1]. You have over-recorded a single side dish by 470 kcal.
The reverse is just as common and more harmful to a goal. Weigh 150 g of raw chicken, cook it, then log the cooked entry at that raw weight, and the error runs the other way. Neither mistake is a rounding error; each is larger than most people's daily deficit. And this is on top of the ordinary drift that makes self-reported intake unreliable in the first place[4].
The rule: weigh raw, log raw
Weigh food in the state the database entry names, and where you have the choice, make that state raw. Raw is the more stable measurement. Cooked weight depends on how long something sat on the heat, how tightly the pot was covered, how much water the rice was given and how well done you like your meat; raw weight depends on none of that.
- Rice, pasta, lentils, burghul: weigh dry, before water.
- Meat, chicken and fish: weigh raw, before the pan.
- Vegetables and fruit: weigh raw, trimmed the way you will eat them.
- Oil: weigh what goes into the pan, not what you imagine ends up on the plate.
- Bread, dairy, packaged food: use the label, which already describes the product as sold.
When raw is not available
Most meals in this part of the world are not assembled from weighed components. Someone else cooked the molokhia, the rice came out of a shared pot, the koshari was bought. In that situation, do not reverse-engineer a raw weight in your head. Log the cooked dish against a cooked entry and accept the wider margin, which is honest, or use the single trick that removes the problem entirely: weigh the whole cooked pot, note what fraction of it you took, and apply that fraction to the raw ingredients you started with.
That works because the raw ingredients are the only thing whose energy content is known with confidence, and dividing a pot into shares needs no nutrition data at all — only a scale and arithmetic.
Consistency beats precision
If you always log rice dry and always log chicken raw, then your record is comparable to itself, week to week. A target adjusted against a consistently biased record still works, because you are moving the number that you are actually measuring. A record that is raw on Monday and cooked on Thursday is comparable to nothing, and the adjustment you make from it is guesswork.
That is also the practical reason the choice is raw rather than cooked: it is the state you can reproduce. For how the target you are comparing against is built in the first place, see the guide to how your daily calories are calculated; for why the protein figure is the one worth getting right when you weigh meat, see the guide to how much protein you actually need.
Sources
- U.S. Department of Agriculture, Agricultural Research Service. FoodData Central, SR Legacy: Rice, white, long-grain, regular, raw, unenriched (169756) and Rice, white, long-grain, regular, unenriched, cooked without salt (169757), 2019. ↩
- U.S. Department of Agriculture, Agricultural Research Service. FoodData Central, SR Legacy: Chicken, broilers or fryers, breast, meat only, raw (171077) and Chicken, broilers or fryers, breast, meat only, cooked, roasted (171477), 2019. ↩
- U.S. Department of Agriculture, Agricultural Research Service. USDA Table of Cooking Yields for Meat and Poultry, Release 2. Nutrient Data Laboratory, Beltsville Human Nutrition Research Center, September 2014. ↩
- Lichtman SW, Pisarska K, Berman ER, Pestone M, Dowling H, Offenbacher E, Weisel H, Heshka S, Matthews DE, Heymsfield SB. Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. New England Journal of Medicine 1992;327(27):1893–1898. ↩