Counting calories without a food scale
You can, to within about 25% per item, if you do three things: measure food against your finger widths instead of guessing by eye, learn the real weight of the five or six foods you eat most often, and reserve the scale for calorie-dense foods. In the trial that tested this, finger-width measurement put 80% of foods within 25% of their true weight, against 29% for cups and spoons.
The rest of this guide is which foods that works on, which foods defeat it, and where the error actually costs you calories.
How accurate is measuring food with your hands?
Good enough to be useful, and better than the measuring cup in your drawer. A 2016 study in the Journal of Nutritional Science had 67 participants estimate 42 pre-weighed foods and liquids, using their finger widths and fists as rulers, then compared the result against household measures.
| Method, geometrically shaped foods | Within 25% of true weight | Within 10% |
|---|---|---|
| Finger width as a ruler | 80% | 13% |
| Cups, spoons and other household measures | 29% | 8% |
Two failures from the same study matter more than the headline:
- Irregular shapes defeat it. Fish fillets, chicken breasts and beef steaks were all estimated more than 50% away from their true weight, by both methods.
- Loose foods defeat the fist. For cereal, mashed potato, muffins and rice, the fist method overestimated, and not a single one of those foods came within 10% of its true weight.
That is worth holding against the hand chart you have seen shared online, the one where a palm is 150 kcal of protein and a thumb is 120 kcal of fat. It is a memory aid, not a measurement, and it was not what the study validated. What the study validated is duller and more useful: hold your fingers next to the food and read off its dimensions.
Which foods do you actually need to weigh?
The ones where being wrong is expensive. Portion error turns into calorie error in proportion to energy density, so the same sloppy 50 g misjudgment costs you almost nothing on vegetables and a whole meal's worth on oil. These are USDA FoodData Central values, with the arithmetic done.
| Food | Kcal per 100 g | Cost of a 50 g misjudgment | Verdict |
|---|---|---|---|
| Olive oil | 884 | 442 kcal | Weigh or count in spoons |
| Peanut butter, smooth | 598 | 299 kcal | Weigh |
| Cheddar, sharp | 410 | 205 kcal | Weigh |
| Chicken breast, roasted, meat only | 165 | 83 kcal | Weigh if you track protein |
| White rice, cooked | 130 | 65 kcal | Eyeball |
| Broccoli, boiled | 35 | 18 kcal | Eyeball |
The working rule: weigh anything above roughly 350 kcal per 100 g, eyeball anything below 150, and use your fingers in between. In a normal week that means the scale comes out for oil, butter, nut butter, cheese, dressings and nuts, and stays in the cupboard for everything else. One tablespoon of olive oil is 13.5 g and about 119 kcal, so a slug poured straight from the bottle is easily 300 kcal you did not see.
What about food you did not cook?
Assume you are underestimating it, because the research says you are. In a 2013 BMJ study of 1,877 adults, 1,178 adolescents and 330 parents interviewed leaving fast food restaurants, two thirds underestimated the calories in the meal they had just bought, and about a quarter were out by at least 500 kcal. Mean underestimation was 175 kcal for adults and 259 kcal for adolescents, and it grew as the meals got bigger.
Three habits fix most of that:
- Look up the chain's published figure before you guess. It is usually higher than your guess.
- Count the sauce, the dressing and the oil the kitchen cooked with as a separate item.
- When there is no figure to look up, record the meal as an estimate and move on. A day with one uncertain meal is still data. A day you skipped entirely is not.
A method that takes thirty seconds a meal
- Photograph the plate before you eat. Whatever you log with afterwards, the photo is the record, and it is the only part you cannot reconstruct later.
- Weigh or count the dense items only. Oil, butter, nut butter, cheese, dressing, nuts. Spoons are fine if the spoon is level.
- Use finger widths for the rest. Read the dimensions off your hand rather than reaching for a cup, which the 2016 trial found to be nearly three times worse.
- Round up when you are unsure. You are correcting a bias the research has already measured.
- Judge the week, not the meal. A consistent method with a known 20% bias shows you the trend. An inconsistent perfect method shows you nothing.
This is also why a photo app and a scale are not rivals. The camera handles the low-density bulk of a plate, where 20% of error is a rounding difference, and the scale handles the fats, where 20% is a meal. If you want the numbers behind what a camera can and cannot see, we wrote them up in how accurate photo calorie counters are.
FAQ
Is the palm and fist method accurate?
Less than the charts suggest. In the 2016 trial, the fist method overestimated amorphous foods such as cereal, mashed potato, muffins and rice, with no food estimated within 10% of its true weight. Measuring a food's dimensions against your finger widths performed far better: 80% of geometrically shaped foods came within 25% of true weight, against 29% using cups and spoons.
Which foods do I really need to weigh?
Anything above roughly 350 kcal per 100 g, because that is where a small misjudgment becomes a large number of calories. Using USDA values, a 50 g error is 442 kcal in olive oil, 299 kcal in peanut butter and 205 kcal in sharp cheddar, but only 65 kcal in cooked white rice and 18 kcal in boiled broccoli.
How many calories is a tablespoon of oil?
One tablespoon of olive oil weighs 13.5 g and contains about 119 kcal, from USDA FoodData Central entry 171413, where olive oil is listed at 884 kcal per 100 g. Cooking oil is the single most common reason an otherwise careful day of logging is wrong by several hundred calories.
Should I round my estimates up or down?
Up. The documented direction of error is underestimation: in the 2013 BMJ study, two thirds of people underestimated what they had just bought and about a quarter were out by at least 500 kcal. Rounding up corrects a bias the research says you already have.
Can a photo app replace the scale entirely?
For low-density foods, in practice yes. For oil, butter, nut butter, cheese and dressings, no, because a photograph cannot show how much fat was absorbed or poured. The workable split is to weigh or count the dense items and photograph everything else.
Does any of this replace professional advice?
No. This is portion estimation, not nutrition advice, and it takes no account of any medical condition, medication or allergy. See the Terms of Service for what Calso is and is not.
Three things to take away
- Finger widths beat cups and spoons by a wide margin, 80% against 29% of foods within 25% of true weight, but both fail on irregular cuts of meat and fish.
- Weigh by calorie density, not by habit. Above 350 kcal per 100 g the scale pays for itself in a single meal. Below 150 it changes nothing.
- Round up, log the uncertain meals anyway, and read the week rather than the plate.
Calso logs the photo for you
Photograph the plate and Calso estimates the calories and macros against targets built from your body and your goal, with no pretence about what a photograph can see. See how it works.
Sources
- Accuracy of hands v. household measures as portion size estimation aids, Journal of Nutritional Science, 2016, 67 participants and 42 pre-weighed foods.
- Consumers' estimation of calorie content at fast food restaurants, BMJ, 2013, 346:f2907.
- Energy values from USDA FoodData Central, SR Legacy entries 171413 olive oil, 172470 peanut butter, 170899 cheddar, 171477 chicken breast, 168878 white rice and 169967 broccoli.