Can You Safely Drink Tap Water Everywhere in Europe? What Compliance Does Not Cover

The honest answer is that nobody can tell you, because nobody has measured the water in your building. European drinking water law is strict and the water leaving a European treatment plant is genuinely excellent. But "compliant supply" is a statement about a supply zone, not about your glass. The pipes between the street and your tap are private property, and in the countries examined here they are never systematically tested. The list of substances that gets measured at all leaves out entire categories, including medicines, antibiotics, hormones and drug residues. Where researchers have gone looking, those substances turn up at nanogram levels and agencies judge the resulting risk very low. Both facts belong in the same sentence, and this article keeps them there while setting out what is checked, what is not, where responsibility legally stops, and how to stop guessing about your own tap.

What a compliance figure actually certifies

Start with the number everyone quotes. In Germany's official 2020 to 2022 report, more than 99 percent of samples met the legal requirements across nearly all microbiological and chemical parameters. That figure is real. It is also narrower than it sounds. It covers 74.1 million people, which is 88.6 percent of the population, in 2,507 supply zones, each of which distributes more than 1,000 cubic metres per day or serves more than 5,000 people.

Two things sit outside that headline. First, self-supply: people on their own source below 10 cubic metres a day, which the German authorities put at roughly 700,000 to 1 million people, have no national figures at all. Second, the sampling is organised per supply zone. Some samples are indeed taken at consumer taps, and that is exactly where the report's lead exceedances turn up, but nothing obliges anyone to sample your building. The health office sets a sampling plan for the zone, not for individual dwellings.

There is also the question of what is on the list. Under the German drinking water ordinance, only pesticides whose occurrence is likely have to be examined. The agency is open about the fact that this makes the figures unrepresentative, and it is explicit about the direction of the bias: because sampling concentrates where contamination is expected, the already low exceedance rates are, in its words, skewed towards higher values and suggest more contamination than is actually present. Fair is fair, and that cuts against the alarmist reading. The narrower point survives intact though: a substance that has no parameter at all, such as a pharmaceutical residue, cannot appear in a compliance statistic in either direction.

Where the utility's responsibility legally stops

This is the part that decides what is in your glass, and it is almost never explained.

In Germany and the Netherlands the formal point of compliance is the tap. But the moment a problem originates inside the building, the law lets go, and in two of these three countries it does not hand the file to anyone else:

• Netherlands. The Drinkwaterbesluit obliges the utility to meet the standards at the delivery point and at the tap, and then states explicitly that this duty does not apply where the cause of the failure lies in the home installation, the collective pipework or a collective water supply.
• Germany. The building's plumbing is a water supply installation in its own right, a Gebäudewasserversorgungsanlage, with its own operator. The ordinance defines that operator as the entrepreneur or other holder of the installation, which in practice is the landlord, the building management or the owner occupier. Not the utility.
• France. The 2022 order transposing the EU rules makes an internal network risk assessment mandatory only in eight listed categories of premises: health establishments, social and medico-social establishments, sheltered housing, crèches and childcare, schools and student accommodation, sports facilities including pools, tourist accommodation including hotels and campsites, and prisons. Every other building, including an ordinary residential block, is merely invited to do it voluntarily, and the order shuts private installations inside dwellings and individual houses out even of that voluntary scheme.
• EU level. The domestic distribution provision of Directive (EU) 2020/2184, in the wording used by the Irish transposition, requires a general analysis of the potential risks associated with domestic distribution systems and states that this general analysis shall not entail an analysis of individual properties. Monitoring may be focused on priority premises.

Read that again: shall not entail an analysis of individual properties. The general analysis certifies the network and checks large public buildings. Your flat is out of its scope, and in France an ordinary apartment block has no mandatory internal risk assessment at all.

Even the one recurring test that does exist inside buildings is aimed elsewhere. German rules require legionella testing only where water is supplied in the course of a commercial or public activity, above a defined tank or pipe volume, where showers or other aerosol producing fittings are present, and the ordinance states plainly that it does not apply in a one or two family house. Where the activity is commercial, which is what letting a flat is, the interval is at least every three years. Where it is public, at least once a year, though the health office may stretch that to three years after three clean years, and may not do so at all in hospitals and care homes. A privately owned family home is tested never, unless the owner pays for it.

The pipes nobody replaced

The strongest evidence that this gap is real comes from the authorities' own replacement programmes.

In the Netherlands, an estimated 233,000 private homes had lead interior pipes when a temporary state subsidy ran from 1999 to 2005, and only around 14,000 owners used it. The national institute notes that it is not known how many of the rest were dealt with by other means such as unsubsidised replacement, renovation or demolition, and that there are probably still many thousands of homes with lead interior pipes.

In Germany, lead pipes had to be removed or taken out of service by 12 January 2026, with two narrow extensions: one where the operator had already commissioned a certified installer who confirms a capacity backlog, and one running to at most 12 January 2036 where the water is used only in the operator's own household and no health risk is of concern. The lead limit itself falls from 0.010 to 0.005 mg/l on 12 January 2028. Deadlines like that exist precisely because the pipes were still there.

Housing stock shows the scale. At the 2011 census, 12.4 million of Germany's 18.9 million residential buildings had been built before 1980. In the Netherlands, 47.4 percent of dwellings dated from before 1975 and 18.1 percent from before 1945 at the start of 2025. Construction year proves nothing on its own about what a specific building's pipes are made of. It does mean that assuming your plumbing is fine because your country's water is fine is an assumption, not a finding.

A brown layer in the glass: what it actually is

If water stands in a glass and leaves a brown or rust coloured deposit, that is not a mystery and it is not nothing.

Chemically, settled brown deposits from drinking water pipes are iron corrosion products: goethite, magnetite and lepidocrocite, plus soluble ferrous phases while wet. The German environment agency describes it in ordinary language: corrosion of unprotected steel leads to brown, unappetising drinking water through oxidised iron compounds, and galvanised steel in hot water lines produces, in its words, mostly just a brown brew from the tap after a few years.

Two consequences follow, and they cut in different directions.

It is not classified as a health limit. Iron at 0.200 mg/l and manganese at 0.050 mg/l are indicator parameters, not health based limits, in Germany and the Netherlands alike. WHO does not propose a guideline value for iron and gives as its reason that it is not of health concern at the levels found in drinking water. So a plumber or a utility telling you discoloured water is an aesthetic issue is quoting the rulebook correctly.

It is still a report on your pipes. Corrosion scale is not inert. Research on distribution systems finds that corrosion scales accumulate heavy metals including lead and arsenic, and that unstable, less protective scale is one of the main factors causing discoloured water. That does not prove you are being exposed to anything, and it must not be read that way. It does mean the deposit is physical evidence that the inside of your pipework is degrading, in a system where nobody is obliged to look.

The UK Drinking Water Inspectorate publishes the one usable marker: where the cause is internal, it is likely the discolouration will persist for many weeks or months. Its advice for discolouration present for less than 24 hours is to run the tap, and to contact the supplier if it lasts longer and does not clear.

In Germany the route is defined by law, and it is worth knowing who owes what. The operator of any water supply installation, meaning the utility for the network and the building operator for the building, must notify the health office without delay of an organoleptically perceptible adverse change such as colour, smell, taste or turbidity, and must then investigate the cause itself. A building operator carries the same duty once the facts suggest the installation is what is changing the water. All of that is triggered by somebody reporting it. Routine zone monitoring is unlikely to find it on its own.

What is not on the list at all

Here is where the standard reassurance stops being an answer.

The German drinking water ordinance's chemical annex contains no parameter for pharmaceuticals, antibiotics or drug residues. There is no limit value and no routine testing programme. WHO states the same at international level: routine monitoring programmes to test drinking water for pharmaceuticals have not been implemented. WHO also gives its reasoning, and honesty requires reporting it: it considers routine monitoring and specialised treatment infrastructure not currently necessary given the limited additional health benefit, and does not consider formal guideline values warranted.

One correction to a claim you will read online: this is not a legal vacuum. German law contains a general obligation that chemical substances must not be present in drinking water at concentrations that give cause for concern about damage to human health, and where the health office becomes aware of such a substance occurring in a supply zone with no limit value in the ordinance, it must set a binding maximum for that zone. There is a legal backstop. What there is not, is anybody routinely measuring.

What happens when researchers do measure:

• The German environment agency reports that at least 414 pharmaceutical active substances, metabolites or transformation products have been detected in the German environment, mostly measured in rivers, streams and lakes, in most cases at up to 0.1 µg/l. On a separate page it names diclofenac, ibuprofen, phenazone, the antibiotic sulfamethoxazole and 17-alpha-ethinylestradiol among substances found in drinking water, at fractions of a microgram per litre, one hundred to one million times below prescribed daily doses.
• A 2025 survey of tap water at 91 German locations found precautionary health related indicator values exceeded for four substances, in one to five samples each, the antibiotic sulfamethoxazole among them by 4 percent above the value and cyanoguanidine the largest at 117 percent. The authors' own conclusion is that with a few exceptions the precautionary values protecting human health are complied with.
• A Dutch survey (RIVM, 2007, sampling in 2005 and 2006) found 15 of 22 target substances in drinking water, most below about 50 ng/l, with salicylic acid and clofibric acid around 125 ng/l. It concluded the amounts were so low that effects on public health are negligible, while noting that treatment considerably lowers concentrations but does not remove all substances completely.
• A French national campaign (ANSES, 2011) of around 280 treated water samples found no quantifiable molecule in 75 percent of them, with more than 90 percent below 25 ng/l cumulative.
• In Brittany, veterinary pharmaceutical residues were found in 20 percent of tap water samples, nearly all below 40 ng/l, with the veterinary antibiotic florfenicol at 159 and 211 ng/l. The authors add the perspective themselves: the detection frequency for every individual compound was below 10 percent, in both raw resources and tap water, and the sites were chosen in watersheds considered sensitive because of intensive livestock farming.

Read those two paragraphs together and you have the actual state of knowledge. Residues are present at nanogram levels. Agencies consistently judge the risk at those levels to be very low. And the same German agency that says effects on human health are not known also says there is no knowledge of what chronic uptake of even low concentrations does. Both halves are true, and only one of them usually gets quoted.

Drugs of abuse: what the measurements really show

Cocaine and other drugs of abuse are measured in European wastewater every year. Whether, and at what level, anything reaches treated drinking water is a separate question with a separate answer.

In wastewater the measurement is systematic and long running: one international programme covers 143 treatment plants in 120 cities across 37 countries. That tells you about consumption, not about your glass.

In treated tap water, a study across eight EU countries analysed 119 samples and found seven of the twelve drugs it looked for, at concentrations from 0.3 to 340 ng/l, in 72 of the samples, which is 60.5 percent. The published abstract does not identify which seven, so no individual substance can be attributed to those samples. The authors quantified what the exposure amounts to: a mean intake through tap water of 0.0064 to 3.531 ng per kg of body weight per day for adults and 0.0247 to 6.7580 ng/kg/day for children. They did not describe this as harmless, and their stated conclusion was that a need clearly exists for further research into the adverse effects on health of inadvertent, sustained exposure to low doses of drugs of abuse. That is the accurate position: the chronic question has not been studied, not that a hazard has been demonstrated.

Treatment removes most but not all of it. In a full scale Spanish waterworks drawing on contaminated surface water, amphetamine type stimulants other than MDMA were removed completely at the pre-chlorination, flocculation and sand filtration stages. Granular activated carbon then removed cocaine completely, MDMA by 88 percent, benzoylecgonine by 72 percent and cotinine by 63 percent, and post-chlorination completely eliminated cocaine and nicotine. Only caffeine and two metabolites, benzoylecgonine and cotinine, persisted all the way through, reduced by 90, 90 and 74 percent respectively. No European regulation obliges anyone to test drinking water for any of these compounds.

France and Germany: potable by law, and still contested

Nothing here says European tap water is undrinkable, and it would be false to claim tap water is forbidden in France or Germany. It is legally potable in both. What is true is more interesting.

A French national campaign run between 2020 and 2022 analysed around 300 sites and changed the national debate. The chlorothalonil metabolite R471811 was quantified in 57 percent of treated water samples and exceeded 0.1 µg/l in 34.1 percent of all treated samples, with a maximum of 2.00 µg/l. Counting all criteria, 45 percent of treated samples failed at least one regulatory threshold. Three caveats belong with that number, and leaving them out would be dishonest: the site selection deliberately combined random sites with sites already suspected of contamination, and non-compliance ran at 58 percent among the suspect sites against roughly 34 percent among the random ones; the 0.1 µg/l figure is a precautionary trace threshold rather than a health limit; and the agency states that no exceedance of a maximum health value was observed anywhere in the campaign.

Then the standard itself moved. In 2024 the same agency reclassified that metabolite as not relevant for drinking water, which shifted its benchmark from 0.1 to 0.9 µg/l. The water did not change. The line did. That is worth sitting with before treating any compliance percentage as a fact of nature.

Official French figures for 2022, covering pesticides and relevant pesticide metabolites, report that 84.6 percent of the population, 56.2 million people, were supplied with permanently compliant water, meaning 15.4 percent, about 10.26 million people, were not at some point in the year, most of them in categories the report itself classes as presenting no health risk. Just over 18,000 people were under actual restrictions. Restrictions are not theoretical either: the health authority for Bourgogne-Franche-Comté maintains a live register of local restrictions, most of them limited to pregnant women and infants on nitrate grounds, alongside a smaller number of general ones, including a restriction in a single commune in force continuously since December 2009.

Public behaviour does not track the compliance rate. Germans drink about 141 litres of bottled water per person per year and the French about 126, against roughly 31 litres in the Netherlands, on industry association figures for 2024. In a 2025 French survey of just over 3,000 people, 86 percent said they were satisfied with their water service, 78 percent said they trust tap water and 68 percent drink it daily, while 75 percent said they were worried about micropollutants including PFAS. High trust and high concern are coexisting, not cancelling out. That is the honest shape of it, and it is why a national percentage settles nothing for one household.

Stop guessing: test your own tap

This is the part we actually believe in. Not fear, not blind trust, evidence about your own building.

1. Get your supply zone report and read it properly. Look at the exceedances and the number of samples, not the annual average, and note which parameters were tested at all. Then remember where it ends: that report stops where your building begins.
2. Have your own tap sampled by an accredited laboratory. The terms of art that actually find one are zugelassene Untersuchungsstelle in Germany, aangewezen laboratorium in the Netherlands and laboratoire agréé in France. A meaningful order asks for lead, copper, nickel and iron plus basic microbiology.
3. Ask for a stagnation sample. Dissolved metals peak in water that has stood in the pipes. German official sampling guidance uses a staged stagnation procedure for lead, copper and nickel, and a four hour stagnation sample is the household approximation of it. The same agency separately advises running off water that has stood for four hours or more before drinking it.
4. Identify your pipes in thirty seconds. Lead is non-magnetic, soft, and scratches to a bright silver that dulls within minutes. Galvanised steel is magnetic and scratches grey. A fridge magnet and a coin settle it faster than any guess about construction year.
5. If you rent in Germany, ask for the lead declaration. Since 13 January 2026 the operator has had to state in text form, and demonstrate, that lead pipes were removed or decommissioned or that the deadline was formally extended. Letting is a commercial activity, so this covers rented housing. Ask in writing.
6. If water is discoloured, time it. Under 24 hours: run the tap, and contact the supplier if it does not clear. Persisting for weeks or months: the cause is likely to be internal, which means your installation.
7. On a private well, test properly and repeatedly. Microbiology and nitrate at minimum, more where there is agriculture or livestock nearby.

Everyday habits that cost nothing: flush the tap until the water runs cold before drinking or filling a kettle, never drink or cook with hot tap water, and do not assume a filter jug is a substitute for knowing what is in the line. Filters address specific, identified problems, they need maintenance to avoid becoming a bacterial habitat themselves, and a filter bought out of vague worry is not evidence. Replacing a lead or corroded pipe is a fix. A jug on the counter is a workaround.

Frequently Asked Questions

Is European tap water safe to drink?

The water in the public network is tested intensively and nearly always meets the legal limits. That statement does not cover the plumbing inside your building, which is legally a separate installation and is not systematically tested, nor the substances that have no parameter at all, including pharmaceuticals, antibiotics and drug residues. Safe is a statement about what was measured.

Does anyone test the water at my own tap?

Almost certainly not. The EU rule on domestic distribution, in the wording used by the Irish transposition, requires a general analysis and states that it shall not entail an analysis of individual properties, with monitoring focused on priority premises such as schools and hospitals. Dutch law suspends the utility's obligation where the cause lies in the home installation, and the French order makes an internal risk assessment mandatory only in eight categories of premises, excluding private dwellings. If you want a measurement of your tap, you commission it.

Why does my tap water leave a brown layer in the glass?

Almost always iron corrosion products from pipes: goethite, magnetite and lepidocrocite. In older buildings the usual source is unprotected or galvanised steel whose zinc layer is spent. Iron is an indicator parameter rather than a health limit, so this is officially an aesthetic matter, but it is also direct evidence that pipework is corroding. Discolouration lasting less than 24 hours usually clears by running the tap; where it persists for weeks or months the cause is likely to be internal.

Are there medicines and antibiotics in drinking water?

Residues are detected at nanogram to sub microgram levels, including diclofenac, ibuprofen, phenazone and the antibiotic sulfamethoxazole. There is no legal limit value and no routine monitoring programme for them in Germany or internationally, although a general obligation exists that substances must not reach concentrations giving cause for health concern, and the health office must set a binding maximum once it becomes aware of such a substance. Agencies judge the risk at measured levels to be very low, and the same agencies state that the effects of chronic exposure at those levels are not known.

Are drugs of abuse present in European water?

In wastewater, unambiguously and measurably, every year. In treated tap water, a study across eight EU countries found seven of the twelve drugs it targeted at 0.3 to 340 ng/l in 60.5 percent of 119 samples; the published abstract does not identify which seven, so no individual substance including cocaine can be attributed to those samples. In the one full scale European treatment plant studied in detail, cocaine was completely eliminated by the end of treatment. No European rule requires drinking water to be tested for these substances.

Should I just drink bottled water instead?

Not automatically. Bottled water is regulated too, and its quality is fixed at the moment of filling rather than tested continuously. Buying bottles also tells you nothing about your building's pipes, which still supply your shower, your kettle and your cooking. A one off test of your own tap answers the question that bottled water only avoids.

Do filters solve this?

Only the problem you have identified. Activated carbon and membrane processes remove a great deal, as the treatment plant data shows, while some compounds persist. A cartridge left in place too long becomes a bacterial habitat. Replacing a lead or corroded pipe is a fix; a jug is a workaround.

The Bottom Line

Europe built an excellent water network and then drew a legal line at the property boundary. On one side of that line water is sampled constantly and reported publicly. On the other side, in the pipes that actually feed your glass, there is no systematic testing, no published data and no obligation on anyone except the building's own operator. That is not a scandal, it is the design. But it means a national compliance percentage is not an answer about your household, and it was never meant to be.

The second gap is the parameter list. Medicines, antibiotics, hormones and drug residues are detected in European water when researchers look, and no routine programme looks. The measured concentrations are low, the agencies judge the risk low, and those same agencies say the long term effects of chronic low level exposure are not known. Anyone who tells you that question is settled, in either direction, is going past the evidence.

So do not outsource it. Read your supply zone report, find out what your own pipes are made of, take a stagnation sample, and get it measured. A brown deposit in a glass is information. Not for tomorrow, and not for someone else to check on your behalf.

To be explicit, because it matters: nothing we sell filters, removes, protects against or offsets anything in your water. Our insulated bottle makes refilling easy, and molecular hydrogen tablets are about dissolved hydrogen gas in water that is already fit to drink, which we set out honestly in what is hydrogen water and what the research actually says. Tap water is not a supplement and a supplement is not water treatment.

Separately, and for reasons that have nothing to do with water quality: dietary minerals come from food, and where a supplement is used the dose should be a known one. Magnesium contributes to normal muscle function and magnesium contributes to a reduction of tiredness and fatigue. These effects relate to dietary magnesium intake generally and not to anything in drinking water, they apply at the daily dose stated on the product, and a varied and balanced diet and a healthy lifestyle remain the basis. We compare the forms in which form of magnesium is best absorbed.

Sources

1. Trinkwasserverordnung (TrinkwV 2023), Bundesministerium der Justiz. Point of compliance at the tap (paragraph 10 no. 1); the building installation as a Gebäudewasserversorgungsanlage with its own operator defined as the entrepreneur or other holder (paragraph 2 no. 2 e and no. 3), with operating duties in paragraph 13; legionella testing limited to commercial or public activity, above defined volumes, with aerosols, and expressly not in a one or two family house, at intervals of three years or one year (paragraph 31); lead pipe removal by 12 January 2026, the two extensions and the operator's duty since 13 January 2026 to declare and demonstrate compliance to the affected consumer (paragraph 17); notification of perceptible changes and the operator's own duty to investigate the cause (paragraphs 47 and 48).
2. TrinkwV paragraph 7, chemical substances. Substances must not be present at concentrations giving cause for concern about damage to health, and where the health office becomes aware of such a substance without a limit value it sets a binding maximum.
3. TrinkwV Anlage 2, chemical parameters. The full parameter list, which contains no pharmaceutical, antibiotic or drug residue parameter; the rule that only pesticides whose occurrence is likely must be examined; and lead at 0.010 mg/l until 11 January 2028 and 0.005 mg/l from 12 January 2028.
4. TrinkwV Anlage 3, indicator parameters. Iron 0.200 mg/l and manganese 0.050 mg/l as indicator parameters rather than health based limits.
5. Drinkwaterbesluit, Netherlands. Article 13(3) requires compliance at the delivery point and at the tap, and article 13(4) states that this duty does not apply where the cause lies in the home installation, collective pipework or collective water supply.
6. Arrêté du 30 décembre 2022 relatif à l'évaluation des risques liés aux installations intérieures, France. Mandatory internal network risk assessment in eight listed categories of premises, a voluntary scheme for other buildings, and the exclusion of private installations inside dwellings and individual houses from that voluntary scheme.
7. S.I. No. 99/2023, European Union (Drinking Water) Regulations 2023, Ireland. A national transposition of Directive (EU) 2020/2184, including the parametric values of Annex I, the exemption threshold of 10 cubic metres per day or 50 persons, and regulation 12 stating that the general analysis shall not entail an analysis of individual properties.
8. Bericht des Bundesministeriums für Gesundheit und des Umweltbundesamtes über die Qualität von Wasser für den menschlichen Gebrauch (Trinkwasser) in Deutschland 2020 bis 2022, Umweltbundesamt, 2025. 74.1 million people, 88.6 percent of the population, in 2,507 reportable supply zones; more than 99 percent sample compliance; 700,000 to 1 million people on self-supply without national figures; and the agency's statement that targeted pesticide sampling biases exceedance rates upwards.
9. Daten zur Trinkwasserqualität, Umweltbundesamt. Lead exceedances found mainly at the consumer's tap as an indication of lead pipes remaining in domestic installations.
10. Rund um das Trinkwasser, Umweltbundesamt guide. Corrosion of unprotected steel producing brown drinking water through oxidised iron compounds, and galvanised steel in hot water lines producing a brown brew after a few years.
11. Trinkwasser-Installation: auf die letzten Meter kommt es an, Umweltbundesamt. Responsibility for the last metres, the advice to run off water that has stood four hours or more, and the escalation route from the building operator to the health office.
12. Brown, black or orange water, Drinking Water Inspectorate, United Kingdom. Causes of discolouration, the advice to run the tap where the problem has been present for less than 24 hours, and the marker that an internal cause is likely to persist for many weeks or months.
13. Physico-chemical characteristics of corrosion scales in old iron pipes, Sarin et al., Water Research, 2001;35(12):2961-9. Composition of iron corrosion deposits including goethite, magnetite and lepidocrocite.
14. Morphological and physicochemical characteristics of iron corrosion scales formed under different water source histories in a drinking water distribution system, Yang et al., Water Research, 2012;46(16):5423-33. Accumulation of heavy metals including lead and arsenic in corrosion scales, and unstable scale as a main factor causing discoloured water.
15. Arzneimittelrückstände in der Umwelt, Umweltbundesamt. At least 414 pharmaceutical active substances, metabolites or transformation products detected in the German environment, mostly in rivers, streams and lakes, in most cases at up to 0.1 µg/l.
16. Umweltwirkungen von Arzneistoffen, Umweltbundesamt. Substances found in drinking water at fractions of a microgram per litre, one hundred to one million times below prescribed doses, alongside the statement that the effects of chronic uptake of even low concentrations are not known.
17. Survey of polar organic micropollutants in German tap waters, Saal et al., International Journal of Hygiene and Environmental Health, 2025;269:114653. Precautionary health related indicator values exceeded for four substances across 91 locations, from 4 percent for sulfamethoxazole to 117 percent for cyanoguanidine, with the conclusion that with a few exceptions the precautionary values are complied with.
18. Geneesmiddelen in drinkwater en drinkwaterbronnen, RIVM, 2007. Detection of 15 of 22 target substances in drinking water, most below about 50 ng/l with salicylic acid and clofibric acid around 125 ng/l, the conclusion that effects on public health are negligible, and the note that treatment does not remove all substances completely.
19. Campagne nationale d'occurrence des résidus de médicaments dans les eaux destinées à la consommation humaine, ANSES, 2011. No quantifiable molecule in 75 percent of treated samples and more than 90 percent below 25 ng/l cumulative.
20. Veterinary pharmaceutical residues in water resources and tap water in an intensive husbandry area in France, Charuaud et al., Science of the Total Environment, 2019;664:605-615. Residues in 20 percent of tap samples, florfenicol at 159 and 211 ng/l, and per compound detection frequencies below 10 percent.
21. Drugs of abuse in tap water from eight European countries: determination by use of supramolecular solvents and tentative evaluation of risks to human health, Muñiz-Bustamante et al., Environment International, 2022;164:107281. Seven of twelve targeted drugs at 0.3 to 340 ng/l in 72 of 119 samples, the estimated intakes, and the authors' call for further research into sustained low dose exposure.
22. Stimulatory drugs of abuse in surface waters and their removal in a conventional drinking water treatment plant, Huerta-Fontela et al., Environmental Science and Technology, 2008;42(18):6809-16. Removal at each treatment stage, complete elimination of cocaine and nicotine by post-chlorination, and persistence of caffeine, benzoylecgonine and cotinine at 90, 90 and 74 percent reduction.
23. Spatio-temporal assessment of illicit drug use at large scale: evidence from 7 years of international wastewater monitoring, Addiction, 2020. 143 treatment plants in 120 cities across 37 countries.
24. Composés émergents dans les eaux destinées à la consommation humaine, ANSES, 2023. The 2020 to 2022 campaign, the chlorothalonil metabolite quantified in 57 percent of treated samples and above 0.1 µg/l in 34.1 percent of all treated samples, the combination of random and suspect sampling sites, and the statement that no maximum health value was exceeded during the campaign.
25. Avis relatif à la pertinence du métabolite R471811, ANSES, 2024. Reclassification of the metabolite as not relevant for drinking water, moving its benchmark from 0.1 to 0.9 µg/l.
26. Les risques liés à la présence de pesticides dans l'eau destinée à la consommation humaine, IGAS, IGEDD and CGAAER, 2024. Official 2022 figures for pesticides and relevant metabolites: 84.6 percent of the population on permanently compliant water and just over 18,000 people under restrictions.
27. Interdictions de consommation de l'eau, Agence Régionale de Santé Bourgogne-Franche-Comté. Live register of local restrictions, most limited to pregnant women and infants, including one in force since December 2009.
28. Loodinname via drinkwater, RIVM, 2019. Approximately 233,000 private homes with lead interior pipes at the time of the 1999 to 2005 subsidy, around 14,000 uses of that subsidy, and the note that many thousands of homes probably still have lead interior pipes.
29. Wohngebäude nach Baujahr, Statistische Ämter des Bundes und der Länder, 2011 census. 12,387,059 of 18,922,618 residential buildings built before 1980.
30. Voorraad woningen naar bouwjaar, CBS. 47.4 percent of Dutch dwellings built before 1975 and 18.1 percent before 1945 at the start of 2025.
31. Bottled water consumption statistics, Natural Mineral Waters Europe, industry association figures for 2024. Germany 141.4 litres per capita, France 126.1 and the Netherlands 31.3.
32. Ce que les Français attendent de leur eau du robinet, Centre d'information sur l'eau and Kantar, 2025 survey of 3,043 people. 86 percent satisfied with the service, 78 percent trusting tap water, 68 percent drinking it daily and 75 percent worried about micropollutants.
33. Pharmaceuticals in drinking-water, World Health Organization, 2012. Routine monitoring programmes for pharmaceuticals have not been implemented, and WHO does not currently consider routine monitoring or formal guideline values necessary.
34. Guidelines for drinking-water quality, 4th edition, World Health Organization. The international reference framework, including the reason given for not establishing a guideline value for iron.

This article is general information about drinking water quality and is not medical advice. It does not claim that any product prevents, treats or protects against any substance in water. If you are concerned about a specific supply or installation, contact your water utility, an accredited laboratory or your local health authority.