Short answer

Short answer

Compare the per-100g column, never the per-serving one. Serving sizes are chosen by each manufacturer and are frequently different between two products sitting next to each other, which makes per-serving figures incomparable.

That is exactly why the per-100g declaration exists. In the UK and EU it is the required basis, and Australian and New Zealand panels carry a per-100g column beside the per-serve one for the same reason: a fixed denominator is the only way two products can be measured against each other.

The serving-size trap

Two granolas sit next to each other. One declares 194 Calories per serving, the other 220. The first looks like the lighter choice.

Then you read the serving sizes: 45g and 50g. Per 100g they are 431 and 440, which is a difference of about 2% rather than the 13% the front figures implied. Nothing was misleading and nothing was hidden. The denominators were just different, and comparing the numerators alone was meaningless.

This runs in both directions, so it is not a matter of catching a trick. A product can look worse than a competitor purely because its manufacturer picked a more realistic serving size. The fix is the same either way: convert both to the same basis before deciding, which the per 100g calculator does, and the per-100g guide explains.

What to compare

Once both are on the same basis, which number matters depends on what you are buying it for.

  • Calories per 100g tells you energy density, which is what matters for foods you eat by volume rather than by a measured amount. The calorie density calculator compares two directly.
  • Protein per 100 Calories is more useful than protein per 100g when you are trying to get protein without the energy that comes with it. A food can be high in protein and high in fat at the same time. Protein per 100 Calories covers the metric, and the protein density calculator works it out.
  • Fibre per 100g is worth a glance on breads, cereals and grain products, where two similar-looking products often differ a lot.
  • Ingredients order is not nutrition data but it is informative. Ingredients are listed by weight, so an item near the front is a large share of the product.

Comparing every field at once is how people end up not deciding. Pick the one that matters for that purchase and ignore the rest.

Front-of-pack claims are regulated, and narrower than they sound

Claims like "light", "reduced fat", "source of fibre" and "high protein" are not marketing language in the EU and UK. They have legal definitions, and a product may only carry them if it meets a specific compositional threshold.

Knowing that cuts both ways. The claim is not a lie, but it usually means something narrower than the impression it creates. "Reduced fat" is a comparison against a similar product rather than a statement that the food is low in fat, and a reduced-fat version of something very fatty can still be fatty.

What "high protein" legally means, and why it surprises people

This is the clearest example of a threshold behaving differently from intuition. Under the EU nutrition claims regulation, a food may be called a source of protein when at least 12% of its energy comes from protein, and high in protein when at least 20% does.

The threshold is a proportion of energy, not an amount of protein. Since protein carries 4 Calories per gram, 20% of energy works out at 5g of protein per 100 Calories, and 12% at 3g per 100 Calories.

That produces results people do not expect. Almonds carry about 21g of protein per 100g, which sounds high, but at roughly 579 Calories per 100g that protein is only about 15% of the energy, so almonds do not meet the high-protein threshold. Skinless cooked chicken breast, at about 31g of protein and 165 Calories per 100g, is around 75%.

Neither fact makes almonds a bad food. It means the claim describes the ratio of protein to energy, which is a different question from how much protein is in the packet, and it is worth knowing which of those two you were asking.

Doing this in a shop without it taking all day

Comparing every product against every other one is not a workable habit. The version that lasts is narrower.

Compare properly once for the products you buy repeatedly, then stop. Your usual bread, yoghurt, cereal and pasta are worth five minutes each, and that decision then holds for months. One-off purchases rarely justify the effort at all.

When you get home, the more useful step is logging what you bought accurately once. The food database entry checker compares a database entry against the packet in front of you, and for anything you eat repeatedly the weighed repeat meal calculator saves the result so it never has to be worked out again.

Limits

A label comparison answers a narrow question about two specific products. It says nothing about which food suits your diet, your budget, your cooking, or how much of it you will eat, and the product you enjoy and use is often the better purchase than the one that wins on a single number.

Declared values are also rounded and carry legal tolerances, so small differences between two products are not meaningful. Treat a gap of a few percent as a tie.

This is general information for adults, not dietary advice, and none of it addresses allergies, intolerances or medical conditions. Speak with a qualified clinician or registered dietitian for individual guidance.

Source note

The protein claim thresholds come from the Annex to Regulation (EC) No 1924/2006 on nutrition and health claims. Label basis rules come from the EU nutrition declaration rules, NHS label guidance and FSANZ nutrition information panels. Composition figures are rounded from USDA FoodData Central.