We analysed the evolution of sales, use and impact of pesticides in EU countries since 2014 to 2024, in the context of changing legislation. This project is a collaboration within the European Data Journalism Network.
- Research, analysis and visualisation: Ana Muñoz Padrós
- Fact-checking: Eva López
- Editing: Gianna-Carina Gruen, Tamsin Walker
-
Food and Agriculture Organization of the United Nations (FAO): Pesticide use per area of cropland: link; Land use: link
-
Eurostat, the Statistical Office of the European Union: Annual sales of active substances contained in plant protection products/pesticides placed on the market from reference year 2011 onwards: link (API)
-
Official Journal of the European Union: Annex III to Commission Regulation (EU) 2021/2010: link
-
Pesticide Action Network InternationalPAN International List of Highly Hazardous Pesticides, December 2024 update: link
-
European Environment Agency (EEA): Percentage of reported water bodies with pesticides exceeding thresholds in rivers in Europe weighted by water body area 2013-2023: link
While sales volume of pesticides don’t account for toxicity and application rate, this indicator is regularly used as a proxy for pesticide use or purchase. Eurostat annually updates sales data as reported by EU countries but monitors pesticide use in agriculture every five years, noting that the comparability across time and territories in this second dataset cannot be ensured. FAO partially relies on sales data to update other indicators. For the EU, data are taken directly from the Eurostat pesticides sales database and mapped to the FAOSTAT pesticides use database.
Eurostat sales data covers "both agricultural and non-agricultural uses” of all active substances “placed on the market in the Member States during the reference period.” Therefore, these data do not account for exports.
Similarly to FAO data, Eurostat pesticide groups are: Fungicides and bactericides; Herbicides, haulm destructors and moss killers; Insecticides and acaricides; Molluscicides; Plant growth regulators; and Other plant protection products.
Eurostat sales data has the following flags:
- Confidential data for: Bulgaria, Hungary, Slovakia, Luxembourg, Estonia, Spain, Finland, Poland, Lithuania, Sweden, Latvia, Malta, Cyprus,
- Breaks in time series: Croatia 2024, Italy 2020
- Definition differs: Some subcategories in 2015 and before in Luxembourg and Bulgaria.
As noted in Eurostat's metadata, from 2016 onwards, geographical and time-series comparability are very good. "Data for the countries which have not yet transmitted reclassified 2011-2015 data according to the revised list of substances are currently flagged ‘d’ in Eurobase". For these countries, from 2011 to 2015, the geographical comparability is lower, and the time-series comparibility is lower as well from 2011 onwards.
When encountering confidential totals for a particular country and year, we have inferred the sum of sales per group as total sales in that country and year.
Eurostat, "for reasons of confidentiality", aggregates the data reported by EU countries before publication in accordance with categories of products and chemical classes indicated in Annex III. As a result, pesticide sales data are disaggregated into three levels and totals, but sales of specific substances are not displayed.
To classify categories and classes, Eurostat uses EU codes; CAS Numbers are not displayed.
Regarding EEA data, "in surface waters, ecotoxicologically-based effect thresholds were assigned to each pesticide (if available) to assess exceedance rates at water bodies.
The sources considered are listed in https://water.europa.eu/freshwater/europe-freshwater/pesticides/documentation/methodology_pesticides_indicator.pdf .
The exceedance of these thresholds indicates pollution by pesticides at levels potentially affecting the aquatic environment. For rivers, effect thresholds could be identified for only 56% of the pesticides reported, so that the actual risk of pesticides in surface waters is likely to be underestimated."
EEA's methodology and annexes for calculating the thresholds can be found here.
The FAOSTAT Pesticides Use domain contains statistics on the agricultural use of major pesticide groups and of relevant chemical families.
Annual data for 2023 were disseminated in June 2025. As of May 2026, the original dataset was last updated on July 15, 2025.
FAO's pesticide data was filtered by 'Use per area of cropland'. We chose this indicator as recommended by expert research, including the Pesticide Atlas published by the Heinrich Böll Stiftung in 2022.
We set a threshold of 10,000,000 hectares of cropland in our countries' selection, based on FAO's Land Use dataset, resulting in a list of 28 non-EU countries and the EU.
Using Eurostat's pesticide sales data, we looked into its evolution alongside policy and political changes. When EU institutions made a move towards reducing pesticides use, did sales drop and vice versa?
We looked into year-to-year changes during the period when a mandatory reduction was in the talks, between 2020 -when the EU Commission announced the proposal- and 2024 -when it was eventually withdrawn-. The chart shows the difference in sales between any given year and the year before. Positive values show a year-to-year increase, negative values mean a decrease.
As most pesticides are sold in Spain, France, Germany, Poland and Italy, we zoomed in on how sales changed specifically in those countries, grouping the rest in one category.
Since the EU proposal had planned for a mandatory 50% cut against a 2015-2017 baseline, this chart shows the increase or decrease in sales per country compared to the average sales of 2015, 2016 and 2017.
We first filtered by total sales per EU country between 2015 and 2024. Then, we calculated the average sales by summing 2015, 2016, 2017 total sales per country and divided by three. This value is our index per country, and we then calculated whether each year from 2018 to 2024 had increased or decreased against this index (in %).
4. Sales in kilograms of organophosphorus herbicides (glufosinate and glyphosate) and highly hazardous pesticides
According to the Annex III to Commission Regulation (EU) 2021/2010 "Harmonised Classification of Substances", level 3 code H99_14 comprises two level 4 codes: H99_14_01 (glufosinate-ammonium) and H99_14_02 (glyphosate). Both substances' CAS numbers are classified as 'Highly Hazardous Pesticides' in PAN List of HHPs, currently based only on classifications by internationally recognised authorities, including international bodies (WHO), European (EU Commission), national agencies (USA EPA, Japan), and the Pesticide Property Database.
Glufosinate-ammonium lost its approval in July 2018. The European Chemicals Agency (ECHA) has signalled health and environmental hazards, making glufosinate ineligible for a six-month grace period in the market.
We have therefore considered that sales of H99_14 from 2019 on refer to glyphosate only. As in the previous chart, we chose to analyse data from 2015 to 2024, since the withdrawn regulation took the 2015-2017 average as a baseline.
Eurostat sales data are not classified by hazardousness. PAN List of HHPs contains all CAS Numbers and names of Highly Hazardous Pesticides. We crossed the CAS Numbers appearing in PAN's list with the list of substances in Annex III, which details the EU codes and CAS Numbers attached to each chemical. The result was a list of the EU codes considered highly hazardous pesticides by official national and international sources.
Since Eurostat sales are aggregated data, we have only considered 'highly hazardous' classes where all substances were classified as such. For example, under the F02_03 class (Dithiocarbamate Fungicides), Annex III has classified six different substances (mancozeb, maneb, metiram, propineb, thiram and ziram). All six are considered HHP according to PAN's research. Our analysis has therefore labelled F02_03 chemical class as HHP.
We chose to analyse pesticide-related pollution in rivers over lakes or groundwater since they tend to flow through populated areas, spread to other water bodies, and cross borders. Aiming to visualise change as in previous charts, we grouped countries in two categories: countries with a higher percentage of rivers exceeding thresholds and countries with a lower percentage. We selected 2018-2023 as the timeframe since 2018 is the first year where all reported countries had available data, and 2023 since it's the most recent reported year.
