Weighing and Tracking Food
Rice triples in weight when you cook it and loses none of its calories. That one fact explains most tracking errors, and the fix takes no extra effort.
Last updated: 30 September 2026
Overview
Cook 100g of rice and you get about 300g of rice. It has not gained calories; it has gained water. Weigh the cooked 300g against the packet's figures and you have just logged three times what you ate.
That is the single most common tracking error, and it is entirely fixable.
Quick Answer
Weigh food raw wherever practical, because packet nutrition labels refer to the raw product. Cooking changes weight by adding or losing water without changing calories, so a cooked weight matched against a raw label can be out by a factor of two or three.
Key Points
- Packet labels are for the raw product. Weigh raw.
- Cooking moves water, not calories.
- If you must weigh cooked, use a cooked-specific database entry.
- Oil is the most under-counted thing in most kitchens.
- Consistency matters more than precision.
Why the numbers move
| Food | 100g raw becomes | Direction |
|---|---|---|
| Rice | 250 to 300g cooked | Absorbs water |
| Pasta | 200 to 240g cooked | Absorbs water |
| Oats | 250 to 300g cooked | Absorbs water |
| Dried lentils | 200 to 250g cooked | Absorbs water |
| Chicken breast | 70 to 75g cooked | Loses water |
| Mince | 70 to 80g cooked | Loses water and fat |
| Bacon | 40 to 50g cooked | Loses a lot of both |
Note the two directions. Weighing carbohydrates cooked makes you over-count; weighing meat cooked makes you under-count. Do both and the errors do not cancel, they just make the total meaningless.
What to do
Weigh raw. It matches the label, it takes no more effort, and it removes the question entirely.
When you cannot — a restaurant, somebody else cooked, a batch that is already made — use a database entry that explicitly says cooked. Most apps have both. "Rice, white, cooked" and "Rice, white, raw" are different entries with very different numbers, and picking the wrong one is the error, not the weighing.
Batch cooking: weigh the whole thing raw, note the total, then weigh the finished batch and divide. If 500g of raw rice becomes 1400g cooked, then 350g of the cooked batch is a quarter of it — which is a quarter of the raw calories.
The bigger leak
Raw versus cooked is a factor-of-three error on one food. Oil is a smaller error on everything, every day, and it adds up faster.
| Item | Actual |
|---|---|
| A "glug" of olive oil | 150 to 300 kcal |
| A tablespoon of peanut butter, heaped | 150 to 200 kcal |
| Oil sprayed for "10 seconds" | 50 to 150 kcal |
| Salad dressing, poured | 100 to 250 kcal |
| Butter on toast, eyeballed | 70 to 150 kcal |
Fat is nine calories a gram against four for protein and carbohydrate, so the things you least expect to weigh are the ones worth weighing most. See how accurate calorie tracking is.
Consistency beats precision
Your true intake is unknowable. Labels carry legal tolerances, your absorption is not identical to the reference, and a database entry is somebody's estimate.
This does not matter, because you are not measuring an absolute. You are measuring a change. If you log the same way every day, then "my intake reads 2,100 and my weight is flat" tells you your maintenance reads 2,100 in your logging — which is the only number you actually need.
The way that breaks is inconsistency: weighing raw on weekdays and guessing at weekends, or logging carefully in week one and loosely in week six. A consistent error is harmless; a drifting one is a plateau you cannot explain.
The habits worth having
1. Weigh, do not estimate, especially fats and calorie-dense foods.
2. Log before eating. After-the-fact logging leaks.
3. Log everything, including the bites while cooking and the drinks.
4. Track all seven days. Five tracked days and two guessed ones is not a week of data.
5. Use verified entries, or build your own from the packet. Crowd-sourced databases are full of wrong numbers.
6. Zero the scale with the bowl on it, then add the food.
When to stop tracking
Tracking is a tool for learning portions, not a life sentence. A few months of it teaches most people what 30g of oats and 150g of chicken look like, and from there an occasional check is usually enough.
If tracking is making you anxious, if you feel unable to eat untracked food, or if the numbers occupy a large part of your thinking, that is worth taking seriously — see not eating enough, and talk to your GP.
Common Mistakes
| Mistake | Why it causes problems | Better approach |
|---|---|---|
| Weighing cooked rice against a raw label | Over-counts by 2 to 3 times | Weigh raw |
| Weighing cooked meat against a raw label | Under-counts by about 25% | Weigh raw |
| Not weighing oil | The largest daily leak in most kitchens | Weigh or spoon it |
| Tracking weekdays only | Two untracked days can erase the week | All seven |
| Picking the first database result | Many are simply wrong | Verified entries |
| Chasing perfect accuracy | It is not achievable | Be consistent instead |
Frequently Asked Questions
Should I weigh food raw or cooked?
Raw, because packet labels refer to the raw product. If you only have it cooked, use a database entry that says cooked.
Why does cooked rice weigh more?
It absorbs water. The calories do not change, so matching a cooked weight to a raw label triples the figure.
Do I need to weigh vegetables?
Not closely. Most are low enough in calories that an estimate is fine. Weigh the calorie-dense things instead.
How accurate does tracking need to be?
Consistent rather than precise. A steady error still tells you what your maintenance reads in your own logging.
Do I have to track forever?
No. A few months teaches you portion sizes, and most people can then check in occasionally rather than log daily.
Sources
British Dietetic Association - https://www.bda.uk.com/
NHS healthy weight guidance - https://www.nhs.uk/live-well/healthy-weight/
NHS food labels guidance - https://www.nhs.uk/live-well/eat-well/food-guidelines-and-food-labels/