Should you weigh food raw or cooked?

This is the question that produces the largest single errors in food tracking, and it has a clean answer with two exceptions.

The answer: raw, when you can

Raw weight is reproducible. A hundred grams of raw rice is a hundred grams of raw rice in every kitchen on earth.

Cooked weight is not. It depends on how much water you added, how long it simmered, whether the lid was on, and how much evaporated. Two cooks making the same dal from the same lentils can finish with cooked weights differing by a third, and their per-100g calorie figures differ by the same proportion.

Composition tables also list most ingredients raw, so raw weight lets you use the reference value directly rather than hunting for a cooked entry that may or may not match your method.

The conversion factors

खाना	मात्रा	कैलोरी	प्रोटीन (ग्राम)	कार्ब्स (ग्राम)	फैट (ग्राम)
White rice	×2.5 to ×3.0	130	2.6	28	0.2
Dried dal and beans	×2.5 to ×3.0	116	7.5	20	0.5
Pasta	×2.2 to ×2.5	158	5.8	31	0.9
Chicken, roasted	×0.70 to ×0.80	165	31	0	3.6
Mutton, braised	×0.65 to ×0.75	258	26	0	17
Spinach, sautéed	×0.25 to ×0.35	45	3	4	2

स्रोत: FAO food energy conversion guidance and USDA FoodData Central

These are the published ranges. Measure your own if you cook the same staple constantly, because a pressure cooker and an open pot do not produce the same yield from the same rice.

The errors this prevents

Cooked rice logged as raw. Weighing 200 g of cooked rice and selecting a raw entry gives 690 kcal instead of 260. A 165 per cent overcount.

Cooked meat logged as raw. Weighing 120 g of grilled chicken and selecting a raw entry gives 139 kcal instead of 198. A 30 per cent undercount, and it also understates protein, which matters for anyone tracking it.

Dry pasta logged as cooked. Weighing 100 g of dry pasta and selecting a cooked entry gives 158 kcal instead of 360.

All three are extremely common and all three come from the same habit: searching a food name and taking the first entry without reading which state it refers to. More on why databases are like this in why two apps give the same roti different calories.

Measuring your own yield factor

Ten minutes, once per staple, and it lasts forever.

Weigh the raw grain. Cook it the way you always cook it. Weigh what comes out. Divide.

If 70 g of raw rice becomes 195 g cooked, your factor is 2.79. From then on you can weigh cooked rice on the plate and divide by 2.79 to recover the raw weight, which gives you the convenience of cooked weighing with the reliability of raw values.

This is the same approach as the household serving calibration in how to track calories without a food scale, applied to a different variable.

The two exceptions

When someone else cooked it. You have no raw weight to work from. Use a cooked entry that says cooked, and accept a wider margin. This covers restaurants, canteens and other people's homes.

When the dish is inseparable. You cannot recover the raw weight of the rice in a biryani. Log the finished dish by weight against a per-100g figure for the mixed dish, as in calories in a plate of biryani.

What yield factors cannot capture

Anything added during cooking. The oil, the ghee, the cream, the sugar.

A yield factor tells you how the weight of an ingredient changed. It says nothing about the 45 g of oil that entered the pot alongside it, which in most home cooking is the largest line in the calculation. That has to be counted separately, always — see how much oil is actually in your curry.

The practical rule, restated

Cook it yourself: weigh raw, log raw, count the fat separately.

Someone else cooked it: weigh cooked, log against an entry that says cooked, assume more fat than you would use.

Never cross the two. That single discipline removes more error from a food log than any other habit on this site.