You weighed every ingredient in a pot of stew and the recipe adds up to 1,840 calories. Dinner is ready. You put 230 grams in a bowl—and suddenly the careful log has no answer.
The missing number is not “six servings.” It is the finished edible weight of the whole cooked batch. Once you have that, the calculation is simple:
total recipe calories ÷ finished batch grams = calories per gram
Then multiply by the grams in your bowl. You can make one portion larger, another smaller and tomorrow's lunch somewhere in between without pretending the pot divided itself into identical sixths.
The worked numbers below are fictional and show the method, not the nutrition of every stew. Checked September 1, 2026 against current USDA, Cronometer and MacroFactor documentation.
One pot, one finished yield
Use a kitchen scale that can safely hold the vessel and stay within its rated capacity. Weigh the empty, cool pot before cooking, or use another safe container whose empty weight you know. After cooking, weigh the vessel and all the edible food together.
Here is the complete worksheet:
| Line | Worked example | | --- | ---: | | Empty pot | 1,240 g | | Pot plus finished food | 2,160 g | | Finished edible batch | 2,160 − 1,240 = 920 g | | Calories from all logged ingredients | 1,840 kcal | | Calories per gram | 1,840 ÷ 920 = 2.00 kcal/g | | Portion in the bowl | 230 g | | Calories in that portion | 230 × 2.00 = 460 kcal |
The result is not a new calorie measurement. It distributes the ingredient total you already logged across the food that actually came out of the pot.
For a different portion, keep the 2.00 kcal/g batch value and multiply again. A 180 g bowl would be 360 calories in this example. A 275 g bowl would be 550. The arithmetic changes with the scale, not with a guessed serving count.
Why the raw ingredient weights no longer divide the pot
Cooking changes weight. Water can evaporate from a sauce or be absorbed by rice and pasta. Fat and moisture can leave meat. The USDA cooking-yield tables measure weight changes from moisture and fat losses under specific cooking methods.
That is why 1,200 grams of ingredients do not promise 1,200 grams of finished food. It is also why copying the calories per 100 g from a raw ingredient onto 100 g of a mixed cooked dish is the wrong operation. The numerator belongs to the whole recipe; the denominator belongs to the finished edible batch.
Cronometer's current cooked-recipe guide uses the same boundary: enter the actual cooked recipe weight, then log a portion in grams. MacroFactor likewise recommends weighing the final dish because cooked food can lose considerable water.
Those app features save arithmetic, but they do not rescue a mismatched ingredient list. The inputs still matter.
Keep the food state consistent
Log each ingredient in the state that matches the amount you used. If you weighed uncooked rice, use an uncooked-rice entry for that weight. If a canned ingredient was drained before it entered the pot, use a drained entry or the applicable label basis. Do not enter a raw weight against a cooked database item and then “correct” it again with the finished yield.
This is especially important when something energy-containing leaves the recipe:
- grease is drained and discarded;
- a marinade or cooking liquid is poured away;
- skin, bones or shells were included on the scale but are not eaten;
- part of an ingredient remains in another bowl or on a baking sheet;
- toppings were counted in the recipe but added to only some portions.
The simple calculation assumes the logged ingredient energy ended up in the edible batch. It cannot infer how many calories left with discarded fat or liquid. When that difference matters, use an appropriate cooked/drained food entry, keep separately added toppings separate, or accept that the result is a practical estimate rather than precision theatre.
Added water is easier. Water changes finished grams but adds no calories. Record it as part of the recipe if the app uses it to model the final weight, then still enter the actual finished yield. A soup that simmers longer can be more calorie-dense per gram even though no extra food was added.
A no-app version you can save
Write these four inputs on one reusable card:
- total calories from the ingredients that actually entered the edible batch;
- empty vessel weight;
- vessel plus finished food weight;
- grams in the portion you are logging.
Then calculate:
finished grams = full vessel − empty vessel
kcal per gram = ingredient kcal ÷ finished grams
portion kcal = kcal per gram × portion grams
Keep more decimal places during the calculation and round only the final portion. In the worked example, 2.00 happens to be exact. If the batch produces 1.8739 kcal/g, rounding that immediately to 1.9 can create avoidable drift across many portions.
Recalculate the next batch when the ingredient amounts or finished yield change. The point is not to create one permanent “my chili” number while the pot gets thicker, leaner or larger every Sunday.
Better logging is still not perfect measurement
Food labels and composition databases already contain variation and rounding. Kitchen scales have limits. Small amounts stay on spoons and pans. A finished-weight calculation makes the portion method internally consistent; it does not turn home cooking into laboratory calorimetry.
That distinction matters if you use intake records with Kcalbit's maintenance calorie calculator. A 14-to-28-day check benefits from a repeatable logging method, but it still needs reasonably complete records and a stable-enough weight trend. One beautifully calculated casserole cannot make an inconsistent month exact.
This is general logging information for adults, not a calorie target, meal plan or instruction to track food. If weighing and logging food causes anxiety, guilt, restriction or compulsive checking, skip the worksheet and seek qualified support. A more consistent estimate is useful only when the process itself is appropriate for you.
For the separate question of why two maintenance tools disagree before food is logged at all, read why calorie calculators give different results. Formula assumptions and cooked-batch math solve different problems.
