The EU's headline food-waste figure counts edible and inedible parts across five sectors, and the household part of it is 69 kilograms per inhabitant. In 2023 the EU generated 58.2 million tonnes of food waste, around 130 kilograms per person, 0.7% more than the 57.8 million tonnes of 2022, with households at 53%, per Eurostat's release of 16 October 2025. The condition that changes what the number means is the denominator: 130 kilograms is a total divided by a population, not an observation of anybody's kitchen.

Where the 130 kilograms sits, sector by sector

Scroll sideways to see the full table →

SectorShare of EU food waste, 2023Per inhabitant
Households53%69 kg
Manufacture of food and beverages19%24 kg
Restaurants and food services11%14 kg
Primary production10%12 kg
Retail and other distribution8%10 kg

Every row comes from the same release (Eurostat, 16 October 2025). The five per-inhabitant figures add up as 69 + 24 + 14 + 12 + 10 = 129 kilograms, while the headline says around 130: rounded components do not recombine exactly, which is worth knowing before anyone builds a further calculation on them.

Three questions that belong beside any food-waste number

What is counted? Here, edible food and inedible parts together — peel, bones and shells included. That is why "130 kilograms of food thrown away" and "130 kilograms of meals thrown away" are different claims.

Which activity, and where? The EU as a whole, with data unavailable for Bulgaria, Spain, Lithuania and Romania and the aggregate estimated using the latest available figure for those countries, as the methodological notes to that release state.

Which year happened, and which year published? The activity is 2023; the release is 16 October 2025. A number quoted as "2025 data" usually means the year of publication. The European Commission's food waste overview states 129 kilograms per inhabitant and over 58 million tonnes, citing Eurostat; where two Commission pages differ by a kilogram, quote the dated release and its 2023 reference year rather than mixing them.

An example: turning the average into a week

Put the statistic through the arithmetic once. Dividing the household share gives 69 kg ÷ 52 = 1.33 kilograms a week per person, so two people account for 2 × 1.33 = 2.66 kilograms a week on the EU average. The whole-chain figure gives 130 kg ÷ 365 = 0.36 kilograms a day.

Both sums are correct and neither describes a household. A population average spreads a measured total across people and days; it does not record what one pair of people did on a Tuesday. The useful direction is the other way: count what your own bin does for seven days, as in a week of less food waste, together, then notice how far it sits from 1.33 kilograms and ask why.

A sector share is also not a moral allocation. "Households are 53%" describes where measured waste arises across the chain; it does not say that households discard 53% of what they buy, and swapping those two sentences is the commonest way this figure is misread.

Municipal waste is a different count again

Municipal waste per person in the EU was 517 kilograms in 2024, 6 kilograms more than the 511 kilograms of 2023, with 248 kilograms per person recycled, or 48.1%, per Eurostat's release of 30 March 2026. The European Environment Agency tracks the same series and puts EU-27 municipal waste at 513 kilograms per capita in 2022 against 503 kilograms in 2010, a 2% rise, with country values from 303 kilograms in Romania to 803 kilograms in Austria.

Do not divide 130 by 517. Food waste is measured across the whole supply chain, including farms and factories; municipal waste is what municipalities collect, and the notes to Eurostat's municipal waste release exclude industrial, agricultural, medical and construction waste from it. The two counts overlap without nesting, and treating one as a percentage of the other produces a number with no referent. What each stream does after collection is a separate question, answered in where does our household waste actually go.

Two readers, two questions, one record

Give each person a question and let them differ. One reads the sector table and asks what retailers and manufacturers are held to; the other reads the household row and asks what a shopping list changes. Write both at the top of the page before either of you opens a comparison.

Then test any new figure against the three questions above. A report that counts only households, or only the edible parts, or a different year, is not comparable to this one, and the honest note is the boundary that differs rather than the number that sounds larger. The same habit applies to any chart with a headline attached, which is the subject of reading a chart before debating its headline.

Sources