Serving size error calculator: Check your food log maths

A serving size error changes your food log in direct proportion to the difference between the amount eaten and the amount entered, when both use the same food and label values. This serving size error calculator multiplies that mass difference by label energy per gram. A negative result means the log is lower; a positive result means it is higher. The result describes a recording mismatch, not a judgement about your meal.

How does the serving size error calculator work?

The serving size error calculator compares label-based energy for your actual portion with label-based energy for your logged portion. Enter energy, its reference mass, the mass eaten and the mass logged. The calculator uses the same energy density for both portions, isolating the effect of the quantity entry.

Start with the interactive calculator above. Its default values are hypothetical: 400 kcal (1,674 kJ) per 100 g (3.53 oz), with 60 g (2.12 oz) eaten and 40 g (1.41 oz) entered. These are teaching numbers, not a tested product or a suggested portion.

The calculation is:

Signed energy error = label energy × (logged mass − actual mass) ÷ label reference mass.

For the default example, the portion eaten corresponds to 240 kcal (1,004 kJ), while the entry corresponds to 160 kcal (669 kJ). The signed difference is −80 kcal (−335 kJ). That is 33.3% below the energy calculated for the actual portion. You can reproduce each result with ordinary multiplication and division.

Energy scales with the serving amount in the FDA label guide, which demonstrates that two servings contain twice the stated energy for one serving. The calculator applies the same relationship to fractional servings and gram quantities. [1]

What is the difference between a serving and a portion?

A serving is the reference amount printed on a food label; a portion is the amount you choose to eat. Your portion can be smaller or larger than the labelled serving. NIDDK distinguishes these two quantities explicitly, so selecting “one serving” in a diary only works when its reference amount matches your portion. [2]

This does not make 170 g the right amount of yoghurt for you. A label serving supplies a reference for arithmetic. The FDA explains that serving sizes reflect amounts typically consumed, rather than a recommendation for how much an individual eats. [1]

How do per-100-g and per-serving entries change the result?

Per-100-g and per-serving entries give equivalent results when the energy and reference mass come from the same column. Mixing energy from one column with the mass from another changes the implied energy density. Check the pair together before comparing actual and logged portions.

FSANZ describes nutrition information panels as reporting average quantities per serving and per 100 g, or per 100 mL for liquids. Energy appears in kilojoules, or in both kilojoules and kilocalories. That makes the column heading as important as the energy number when you copy an Australian label. [3]

Here are calculated examples using invented labels. Energy figures are rounded to whole kcal and kJ; mass conversions are approximate. None of these rows estimates how often people make an error.

A whole-package column creates another possible reference. CDC explains that dual-column US labels can show both per-serving and whole-package information. If you copy whole-package energy, pair it with whole-package mass rather than the smaller serving mass. [4]

Can you mix grams, ounces, kilograms and pounds?

Grams, ounces, kilograms and pounds can be compared after conversion to the same mass unit. Millilitres, cups and fluid ounces measure volume, so converting them to food mass requires additional information. This calculator accepts mass inputs only and displays energy in both kcal and kJ.

The calculator uses 1 kcal = 4.184 kJ, 1 oz ≈ 28.3495 g and 1 lb ≈ 453.5924 g, consistent with NIST conversion tables. Ounces here mean ordinary weight ounces, not fluid ounces. [5]

Suppose a hypothetical prepared dish has 120 kcal (502 kJ) per 100 g. The portion eaten is 0.25 kg (0.551 lb), while the log represents 0.20 kg (0.441 lb). Converted to grams, those are 250 g and 200 g. The calculated energy is 300 kcal (1,255 kJ) eaten versus 240 kcal (1,004 kJ) logged: a −60 kcal (−251 kJ) difference.

Changing a unit selector converts the existing value, preserving the quantity. You can also enter each mass in a different unit. That makes a gram-based label compatible with a scale displaying ounces without treating “6 oz” as “6 g”.

For a volume-only label, look for a product-specific mass equivalent before using this tool. A per-100-mL drink entry cannot be paired directly with a gram weight without knowing the relationship between those quantities. Keeping volume and mass separate avoids adding an unsupported assumption to otherwise simple arithmetic. [3][5]

Does a 20% portion error mean a 20% calorie error?

A 20% portion-mass error produces a 20% energy error when the same energy density applies and the actual portion is the denominator. The percentage describes this one entry. It does not establish a 20% error across your whole day or predict a change in body weight.

Consider an invented label with 200 kcal (837 kJ) per 100 g. If you eat 100 g (3.53 oz) and enter 80 g (2.82 oz), the log contains 160 kcal (669 kJ). It is 40 kcal (167 kJ), or 20%, below the actual portion estimate.

The denominator matters. Moving the entry from 160 to 200 kcal is a 25% increase relative to the old entry, even though the original error was 20% relative to the actual portion. Neither percentage contradicts the other. They answer different questions.

Small mass differences also have different energy consequences across labels. A hypothetical 10 g (0.35 oz) mismatch at 1 kcal per gram creates a 10 kcal (42 kJ) difference. At 6 kcal per gram, the same mismatch creates 60 kcal (251 kJ). These are mathematical examples, not nutrient values assigned to particular foods.

How reliable is the actual portion estimate?

The actual portion estimate is only as reliable as the measurement or information you enter. The calculator can quantify a known difference, but it cannot discover an unknown portion. Research on portion estimation supports treating remembered or visually selected amounts as estimates rather than exact measurements.

A 1994 study compared photo-assisted portion estimates with 14-day weighed records in 144 adults aged 40 to 64. Between 85% and 95% chose the closest photo or a neighbouring photo, yet relationships with measured portions were relatively weak for most foods. Statistically significant correlations appeared for only three of eight foods studied. [6]

That study assessed a food-frequency questionnaire with four photo options per food. It did not test kcalm, modern phone cameras or AI food recognition. Its results offer context for uncertainty, not an accuracy score for an app. [6]

If your “actual” amount is another visual guess, the output compares two guesses. You can still use that comparison to explore how much the entry changes, but it is more honest to call it a scenario. For measurement choices, see our food scale, measuring cups and portion-error guide.

How can you correct a serving-size mismatch in your food log?

You can correct a serving-size mismatch by checking the food, matching the label reference and updating the quantity represented by the original entry. Keep the correction tied to the same meal. The purpose is a clearer record of what you ate, with a repeatable process for future entries.

1. Check the product and preparation description against the food you ate.
2. Read the energy value and its reference mass from the same label column.
3. Establish the amount eaten, using a measurement or a clearly identified estimate.
4. Compare that amount with the mass represented by the current diary entry.
5. Update the original quantity and review the recalculated energy.

For the yoghurt example, the correction is from one 170 g serving to 1.5 servings, or directly to 255 g if your diary accepts grams. Adding another full serving would instead represent 340 g (11.99 oz), overshooting the known portion. The smaller, targeted edit makes the record easier to interpret later.

This calculator assumes the food composition is the same for both quantities. If one entry describes a dry ingredient and the other a prepared meal, that assumption needs attention first. Our raw versus cooked weighing guide discusses that separate matching problem. For recipe-level questions, see homemade meal calorie maths.

When reviewing a meal in kcalm, this calculation can serve as an external cross-check. It does not measure kcalm accuracy or replace checking the food description. If a small difference is uncertain, consider recording that uncertainty rather than repeatedly adjusting a number you cannot verify.

Frequently Asked Questions

These frequently asked questions explain the serving size error calculator and its limits. The examples use label-based energy estimates, with rounded unit conversions, and do not prescribe portions or daily calorie targets.

What does a negative result mean?

A negative result means the logged energy is lower than the energy calculated for the actual portion. For example, −80 kcal (−335 kJ) means the entry is that much lower under the same label assumptions. A positive result means it is higher.

Can I enter more than one serving?

Yes. Multiply the serving mass by the number of servings represented. If one serving weighs 50 g (1.76 oz), two servings weigh 100 g (3.53 oz). Use those masses with the energy and mass for the original label serving.

What happens if I leave a field blank or enter zero?

Blank fields produce a request for a value. Reference mass is positive because it is a divisor. Actual and logged masses can be zero, and zero label energy is accepted. Percentage error is unavailable when actual label-based energy is zero, although the absolute difference remains calculable.

Does this tell me the exact energy I absorbed?

No. It calculates energy from the label values you supply. FSANZ describes panel values as averages. Correcting a portion entry does not turn those averages into a direct measurement of your individual food or its digestion. [3]

Can the result tell me how much less to eat tomorrow?

No. The result is a difference between two representations of the same portion. It supplies no personal intake recommendation or compensation rule. You can correct a record without turning the edit into a new eating target.

Sources

These sources were retrieved on 2 November 2026. Government guidance supports label interpretation, NIST supplies unit conversions, and the research abstract supplies the portion-estimation context. All calculator defaults and worked food-label examples are hypothetical arithmetic.

1. FDA: How to understand and use the Nutrition Facts label.
2. NIDDK: Food portions, choosing just enough for you.
3. Food Standards Australia New Zealand: Nutrition information panels.
4. CDC: Nutrition Facts label and your health.
5. NIST: Conversion factors, units listed alphabetically.
6. Haraldsdóttir and colleagues: Validity of individual portion size estimates in a food frequency questionnaire, 1994.