Skip to content
A ranking of cannabis in wastewater is doing the rounds on social media: Erfurt supposedly has the third-highest cannabis levels of any city in Europe, with four German cities in the top six. We checked the figures against the raw data. They are correct. What the measurement can and cannot tell you, however, gets lost in the hype. We set out both here so you can explain it properly to members, neighbours or the local press.
What the ranking claims and what is correct
The basis is the European wastewater study by the EU drugs agency EUDA and the SCORE research network, which measures cannabis and other substances in wastewater every year. If you sort the cannabis values of the European cities in the 2025 raw data, the top of the list looks like this (in mg THC-COOH per 1,000 inhabitants per day): Geneva 252.8, Amsterdam 194.1, Erfurt 178.5, Berlin 172.7, Hamburg 167.3, Nuremberg 159.0 (EUDA and SCORE, 2026b). Four German cities in the top six, Erfurt in third place: that part is accurate. At the very top of the raw data, though, are Montreal, Seattle and Québec. The EUDA lists cities outside Europe only as a global comparison.
Three points get lost along the way.
First, there is no official ranking. The EUDA publishes values per city and describes patterns, for example that the highest loads were measured in cities in the Netherlands, Germany and Slovenia (EUDA and SCORE, 2026a). The ranking only appears once someone sorts the table.
Second, a value of 187 is also circulating for Erfurt. The raw data say 178.51. If you quote it, use the CSV as your source, not an article about the CSV.
Third, Geneva is not part of the European analysis. For 2025 the EUDA counts cities in 25 countries: the EU, Norway and Türkiye (EUDA and SCORE, 2026a). Switzerland appears in the raw data with twelve cities, Geneva being the highest. In the text of the study, Swiss cities only show up in the global comparison, alongside Canada and Chile. Within the European analysis, Erfurt would be in second place behind Amsterdam. This too shows how much a ranking depends on who puts the list together.
How cannabis in wastewater is measured
For one week, the participating treatment plants take a daily 24-hour composite sample from the inflow; in 2025, most cities sampled during a single week between March and May. The labs measure THC-COOH, the main breakdown product of THC. Concentration, flow and the population served give the load per 1,000 inhabitants per day. In 2025, 115 cities from 25 countries took part (EUDA and SCORE, 2026a), with around 72 million people connected to their treatment plants in total (our own sum from the site table, EUDA and SCORE, 2026d).
The advantage over surveys: nobody has to answer honestly. All labs follow a common protocol and take part in interlaboratory comparison tests. That is why the values are comparable between SCORE cities, and that is exactly what the study is built for.
What the number does not tell you
The EUDA itself states that the method provides no information on the prevalence and frequency of use, on the groups of users or on the purity of the substances (EUDA and SCORE, 2026a). For cannabis in wastewater, four further limits apply.
The load is not a head count. A city with a few daily users can have the same value as one with many occasional users. Higher THC content in products also raises the value, without anyone using more.
The excretion factor is uncertain. Only a small share of the THC taken in ends up as THC-COOH in urine; assumptions in the literature range from about 0.5 to 2.5%. Converting the load into grams of cannabis per person is therefore not reliable. The EUDA deliberately does not convert.
A large part sticks to solids. THC-COOH is fat-soluble. In a lab experiment with stool samples from nine users, only about 40% of it passed into the water; the rest stayed on the particles (Pandopulos et al., 2022). Filtration and pH adjustment are considered the most critical steps in the lab (Causanilles et al., 2017). Within SCORE this is standardised; between different monitoring programmes it is not.
One week is one week. Weather, public holidays, events and commuter flows shift the result. The population served by a treatment plant rarely matches the city boundary: in Dresden, around 672,000 people are connected to the Kaditz plant, some of them in the surrounding area. Tourist cities like Amsterdam have far more people in the network during the day than are registered there.
What the data do show fairly well is the pattern across the week. The EUDA writes that cannabis, unlike cocaine or MDMA, is spread fairly evenly over the week (EUDA and SCORE, 2026a). In Dresden, for example, weekdays (119.9) and weekends (110.7) are close together (EUDA and SCORE, 2026b). We read this as a better fit for regular use than for a weekend phenomenon. That is our interpretation, not the study’s.
Why there is no year-on-year comparison
In the EUDA change data, all 13 German cannabis rows carry the note “no previous data” (EUDA and SCORE, 2026c). In the years 2022 to 2024, German labs did report cocaine, amphetamine or MDMA to SCORE, but no cannabis values. The most recent German cannabis values in the EUDA data date from 2013, from Dortmund, Dresden and Dülmen. Twelve years in which population estimates, sampling and labs may have changed are no basis for comparison. So the European study cannot tell you whether anything has changed since the German Consumer Cannabis Act (KCanG) came into force. There is a German baseline measurement all the same, from a separate project at TU Dresden. This is the subject of a separate post.
What this means for your club
When the ranking makes the rounds in your city, questions follow. Three statements hold up in almost every conversation.
The number describes the amount a whole city excretes in one week. It does not say how many people use cannabis, let alone whether they get it from a club, their own garden or the black market.
Rankings are made by whoever sorts the table. The measurement is designed for comparisons between cities, but a ranking without the limits next to it is a headline, not an analysis.
A high value is no proof that the KCanG has failed, and a low one is no proof of its success. The European study simply has no time series for Germany that could show either.
For prevention talks at the club, the study is still useful. It shows that cannabis is an everyday topic in German cities and not a fringe phenomenon, which is why prevention matters, and it works well for explaining to members why research works with data from wastewater rather than data about individual people. In Saxony and Thuringia, crystal meth belongs in these conversations too, see Cannabis and Meth in Wastewater: What Chemnitz, Dresden and Erfurt Show.
Key takeaways
- The values in the circulating ranking match the EUDA and SCORE raw data if you sort only the European cities. The EUDA does not publish an official ranking.
- For Erfurt, the raw data say 178.51, not 187.
- Wastewater measures the total amount of a breakdown product, not a number of people, not frequency and not potency.
- The excretion factor, binding to solids and the single measurement week make conversions into grams unreliable.
- The European study has no previous-year values for Germany. No conclusions about the KCanG can be drawn from it.
Sources
- Causanilles, A., et al. (2017). Improving wastewater-based epidemiology to estimate cannabis use: focus on the initial aspects of the analytical procedure. Analytica Chimica Acta. doi:10.1016/j.aca.2017.08.011
- EUDA and SCORE (2026a). Wastewater analysis and drugs: a European multi-city study. Updated 18 March 2026. https://www.euda.europa.eu/publications/pods/waste-water-analysis_en
- EUDA and SCORE (2026b). Raw data of the 2025 wastewater study (ww2026-all-data). https://www.euda.europa.eu/sites/default/files/data/data-nodes/33337/versions/5/ww2026-all-data_en.csv
- EUDA and SCORE (2026c). Changes compared with the previous year (ww2026-changes-all-substance). https://www.euda.europa.eu/sites/default/files/data/data-nodes/33337/versions/5/ww2026-changes-all-substance_en.csv
- EUDA and SCORE (2026d). Data package of the 2025 wastewater study, site table with connected population (ww2026-site-info-table). https://www.euda.europa.eu/data/repository/drugs-municipal-wastewater-europe-source-data-2026_en
- Pandopulos, A. J., et al. (2022). Partitioning of phytocannabinoids between faeces and water: implications for wastewater-based epidemiology. Science of the Total Environment, 805. doi:10.1016/j.scitotenv.2021.150269
