France's food hygiene inspection results: the Alim'confiance barometer by département
Alim'confiance publishes the result of the latest official hygiene inspection of every food business inspected in France. Of the 71,902 results published for inspections carried out from September 27, 2025 to September 25, 2026, 93.8% are rated "very satisfactory" or "satisfactory", 5.7% "needs improvement" and 0.5% "urgent correction required". When the inspection follows a complaint, a foodborne outbreak or a targeting decision, the share of the two unfavourable ratings reaches 18.6%, against 6.6% for scheduled inspections.
Extracted on September 26, 2026. For the result of a particular establishment, see the official site alimconfiance.agriculture.gouv.fr.
The four hygiene levels, nationally
71,902 published results. Each level is defined in French regulation (Article D. 231-3-11 of the Code rural); the ratings describe the establishment's hygiene as a whole on the day of inspection, not any particular point.
Very satisfactory (Très satisfaisant): 24.5% (17,631)
No non-compliance, or only minor non-compliance.
Satisfactory (Satisfaisant): 69.3% (49,832)
Non-compliance that does not justify administrative enforcement measures; the authority sends a letter reminding the operator of the rules so that practices improve.
Needs improvement (À améliorer): 5.7% (4,110)
The operator has been served formal notice to take corrective measures within a deadline set by the authority, which leads to a new inspection by state services to check they are in place.
Urgent correction required (À corriger de manière urgente): 0.5% (329)
Non-compliance liable to endanger consumer health; the authority orders administrative closure, or the withdrawal or suspension of the health approval.
Types follow the Alim'confiance site's own filter. Among them, the share of the two unfavourable ratings ranges from 3.5% (supermarkets and grocery stores) to 8.1% (food trades). Activities are not inspected at the same rate nor for the same reasons: the gap describes the published results, not a hygiene ranking of trades.
Hygiene levels by type of establishment
Type
Results
Very satisfactory (%)
Satisfactory (%)
Needs improvement (%)
Urgent correction required (%)
Restaurants
33,070
17.8
75.7
6.0
0.5
Food trades (butchers, bakers, etc.)
14,389
14.6
77.3
7.4
0.7
Other (processing, slaughter, transport, storage, etc.)
9,496
48.4
47.6
3.8
0.3
Institutional catering (canteens, central kitchens)
7,024
28.9
65.2
5.8
0.1
Supermarkets and grocery stores
6,549
36.1
60.4
3.1
0.4
Farm producers
1,374
46.7
45.9
7.0
0.4
By reason for inspection
This is the reading that prevents a misreading. An inspection triggered by a complaint, an outbreak (TIAC) or targeting goes to an establishment where a problem is already suspected, and its results are clearly less favourable. Follow-up inspections, which a "needs improvement" rating entails by definition, end in "satisfactory" in 98.3% of published results; since only the latest result is shown, they often replace the result that caused them.
Hygiene levels by reason for inspection
Reason
Results
Very satisfactory (%)
Satisfactory (%)
Needs improvement (%)
Urgent correction required (%)
Scheduled inspection
53,917
30.0
63.3
6.4
0.3
Follow-up inspection
13,679
0.1
98.3
1.1
0.5
Targeted inspection
1,588
23.4
58.1
14.0
4.5
Complaint or report
1,294
21.7
60.0
15.9
2.4
Approval request
908
72.2
27.1
0.3
0.3
Foodborne outbreak (TIAC)
516
21.7
58.5
18.6
1.2
By region and département
The département is that of the establishment's commune (INSEE code). Among départements with at least 100 results, the share of the two unfavourable ratings ranges from 1.1% to 32.8%, with a median of 6.2%. These gaps also reflect each area's mix of establishments, inspection programme and reasons for inspection: they do not rank the hygiene of départements.
Below 30 results only the count is shown: a breakdown over so few inspections would say nothing stable and could point to one establishment. Open a region to see its départements.
This table is not a trend. Only the latest result of each establishment stays published: an older "needs improvement" result has often been replaced by the follow-up inspection already, whereas a recent one is still waiting for its own. That mechanism alone pushes the unfavourable share up towards recent months, and it cannot be read as a deterioration. The last three weeks are also incomplete, and not at random: in the extract, "needs improvement" and "urgent correction required" ratings stop appearing about three weeks before the extraction date, whereas favourable ratings show up within days (the official FAQ gives the operator 15 days to submit observations before publication). The unfavourable share of the last month is therefore understated. The first and last months are partial.
Published results by month of inspection
Month
Published results
Needs improvement or urgent
September 2025 (partial)
685
3.2%
October 2025
7,258
2.6%
November 2025
6,094
2.9%
December 2025
5,707
3.0%
January 2026
5,654
2.8%
February 2026
5,485
3.4%
March 2026
7,033
5.0%
April 2026
6,620
5.9%
May 2026
5,362
7.2%
June 2026
8,076
8.9%
July 2026
6,722
12.8%
August 2026
4,931
15.2%
September 2026 (partial)
2,275
3.3%
Method and limits
Source: the dataset "Alim'confiance, résultats des contrôles officiels sanitaires" published by the French Ministry of Agriculture (Ministère de l'Agriculture, de l'agro-alimentaire et de la souveraineté alimentaire), under the Licence Ouverte (Etalab). Extracted on September 26, 2026; aggregated with group-by queries on the public API, without downloading the named rows. dgal.opendatasoft.com
Coverage: according to the official FAQ, the result published is that of the establishment's latest inspection, and the data stay accessible for one year. The extract holds inspections from September 27, 2025 to September 25, 2026. Each row is therefore an establishment's latest known result over a rolling twelve months, not every inspection of the year. FAQ Alim'confiance
Publication lag: unfavourable results from the last three weeks are not yet in the extract, so the national shares slightly understate them. Separately, the public Alim'confiance website was unavailable from spring 2024 to 26 March 2026; inspections kept being recorded during that period, and the open dataset covers every month of our window.
Not a random sample: official controls are carried out on a risk basis and with appropriate frequency (Regulation (EU) 2017/625, Article 9), or triggered by a complaint, an outbreak or a follow-up. The shares describe the published results, not the situation of every establishment in France. EUR-Lex
What the data do not say: which points motivated a rating. The level describes the establishment's hygiene as a whole; no particular topic (cleaning, premises, traceability, temperature control or anything else) can be inferred from it, and no cause is put forward here.
Processing: 1 row without a level is excluded (71,903 rows in all). The département comes from the establishment's INSEE commune code; 1,794 rows without a geolocated département were placed that way, 353 rows whose geolocated département differs from the commune's follow the commune, and 31 rows without a usable commune code are in no département (they stay in the national total). No name, SIRET or address is kept.
How to cite this barometer
KelvinTrack, "Alim'confiance food inspection barometer", based on Alim'confiance data (French Ministry of Agriculture, Licence Ouverte (Etalab)), extracted on September 26, 2026. https://kelvintrack.com/en/shop/resources/food-inspection-results-france
The source data are under the Licence Ouverte (Etalab open licence): reuse is free provided the source and date are credited. The aggregates on this page can be downloaded: CSV: France, regions, départements
