Is lithium in drinking water dangerous?

At the levels found in US drinking water, there is no evidence it is harmful. Lithium is a naturally occurring mineral that dissolves into groundwater. Psychiatric lithium is dosed at 600 to 1,200 mg per day of a lithium salt; drinking water even at a high 100 µg/L delivers roughly 0.25 mg per day — about 1,000 to 4,000 times below a therapeutic dose. This is why the EPA set only a non-binding Health Reference Level of 10 µg/L rather than an enforceable limit. Documented harms — kidney and neurological effects — occur at pharmacological doses, not at water levels.

Why was lithium the most-detected substance in EPA's UCMR 5?

In EPA's Fifth Unregulated Contaminant Monitoring Rule (2023-2025), lithium was detected in 36.4% of 10,299 public water systems — more than any other analyte on the list — with a median detected concentration of 23.3 µg/L and a maximum of 960 µg/L. It is common because it is naturally abundant in geology and dissolves readily into groundwater, especially in arid Western-US aquifers over lithium-bearing formations. UCMR 5 is a screening program for unregulated contaminants, so a high detection frequency reflects how widespread the element is, not that it is unsafe.

Is the lithium in water the same as psychiatric lithium?

It is the same element, but the doses are on completely different scales. Prescription lithium (lithium carbonate) is given at 600 to 1,200 mg per day and produces measurable effects on mood, kidney function, and thyroid at those levels. Water-borne lithium exists as trace amounts of the Li+ cation. Across 1,464 US public-supply wells, EPA estimated a maximum daily ingestion of 0.05 to 1.0 mg from drinking 2.5 liters per day — versus a low therapeutic dose of 600 mg. The dose gap is the whole story.

How do I remove lithium from water?

Lithium cannot be removed by boiling, heating, or disinfecting the water. According to EPA, ion exchange is effective, and reverse osmosis reduces lithium by roughly 80 to 95 percent or more — so a point-of-use RO system is the straightforward choice if a well or system reports high values and you want to reduce them. A standard activated-carbon filter does little, because lithium is a dissolved monovalent cation that carbon does not adsorb well.

Lithium in Drinking Water: Is It Safe?

At a glance

Chemical family	Alkali metal (Group 1) — the lightest metal, naturally occurring and mobilized from geology into groundwater. In water it exists as the Li⁺ cation. It is not a PFAS; it appears on the UCMR 5 list because that rule bundles a set of priority unregulated contaminants for monitoring.
CAS number	7439-93-2 (PubChem CID 3028194)
Classification	No EPA Maximum Contaminant Level; no EPA Health Advisory. EPA screening Health Reference Level (HRL) 10 µg/L (from a provisional oral reference dose of 2 µg/kg-day); UCMR 5 minimum reporting level (MRL) 9 µg/L (EPA Method 200.7). Not classified by IARC. Note: lithium carbonate — the pharmaceutical salt, not water-borne Li⁺ — is on California Proposition 65 for developmental toxicity (listed 3/19/1999). No WHO, Health Canada, Australian, EU, or UK drinking-water limit.
Where you encounter it	Groundwater is dominant, especially arid/Western-US aquifers over lithium-bearing geology — the reason it is the most-detected UCMR 5 analyte. Anthropogenic sources are secondary: lithium-ion battery manufacture and disposal, mining and extraction, catalysts, and pool/spa sanitizers. Diet also contributes (grains, leafy greens), and whether trace lithium is a beneficial micronutrient is an open debate.
Sleep micro environment relevance	Indirect and speculative. Pharmacological lithium modifies circadian period and sleep architecture — but at drug-dose serum levels, not at trace water levels ~1,000× lower. On the Atlas as part of the tap-water picture; the sleep link is mechanistic speculation only.
Activated carbon capture	Not removed by a standard carbon filter. Per EPA, ion exchange is effective and reverse osmosis reduces lithium ~80–95%+, while activated carbon does little because Li⁺ is a dissolved monovalent cation. Not removed by boiling, heating, or disinfection. Regulatory — US EPA

Regulatory & certification status

Lithium is a genuinely unregulated contaminant: no country has set an enforceable drinking-water limit, and the US numbers are screening values only. The rows below give the honest state of play — including where the answer is simply "not a regulated parameter."

United States	EPA has not established a Health Advisory or regulatory standard for lithium. It publishes a non-binding Health Reference Level of 10 µg/L (derived from a provisional oral RfD of 2 µg/kg-day) as a screening benchmark, and set a UCMR 5 minimum reporting level of 9 µg/L. Regulatory — US EPA
US — California Prop 65	Lithium carbonate (the pharmaceutical salt, not the trace Li⁺ in water) is listed for developmental toxicity, effective 3/19/1999. This is a drug-dose listing and does not translate to a water-level hazard. Regulatory — California OEHHA
European Union	Lithium is not a regulated parameter in the EU Drinking Water Directive (2020/2184) — no limit exists. Regulatory — honest absence
Canada	The Guidelines for Canadian Drinking Water Quality contain no guideline value for lithium. Regulatory — Health Canada
Australia & WHO	The NHMRC Australian Drinking Water Guidelines set no health-based value for lithium, and the WHO Guidelines for Drinking-water Quality have no guideline value for it either. Regulatory — NHMRC
The 72-hour test window	Not applicable. Lithium is a waterborne ingestion exposure measured by a standard water-lab test, unrelated to the VOC-emission chamber tests used for mattresses and foam. Inferred — from the ingestion-via-water route versus the material/VOC focus of product emissions testing

What it is

Lithium is the lightest metal and the third element on the periodic table — a soft, reactive alkali metal that never occurs free in nature. In water it appears only as the Li⁺ cation, dissolved out of lithium-bearing rock. Despite its high-profile modern uses in batteries, it is at heart a common mineral in the earth's crust, and its presence in groundwater is overwhelmingly natural.

The single most important fact about lithium in water is the dose gap. Psychiatric lithium — the mood-stabilizing drug — is prescribed at 600 to 1,200 mg per day of a lithium salt, doses at which it measurably affects mood, kidney function, and the thyroid. Drinking water, even at a high 100 µg/L and 2.5 liters per day, delivers about 0.25 mg per day — roughly 1,000 to 4,000 times below a therapeutic dose. EPA made the same point with its own occurrence data: across 1,464 public-supply wells (the 21–396 µg/L range from the third quartile to the maximum), drinking 2.5 L/day yields a maximum daily ingestion of 0.05 to 1.0 mg, against a therapeutic low end of 600 mg. Regulatory

Because it is on the UCMR 5 list, lithium gets grouped in headlines with genuinely worrying unregulated contaminants like PFAS. It is worth being precise: lithium is not a PFAS, not a synthetic industrial chemical, and not a carcinogen — IARC has never issued a monograph on it. It is a naturally abundant mineral that happens to be very common in groundwater.

Where you encounter it

From geology and groundwater (dominant)

Lithium's presence in drinking water is overwhelmingly a groundwater story. EPA notes it runs higher in arid Western-US groundwater sitting over lithium-bearing formations — which is exactly why it topped the UCMR 5 detection list. This is the same pattern as other naturally-occurring well constituents: the geology, not human activity, sets the baseline.

From human activity (secondary)

Anthropogenic sources exist but are secondary to geology: lithium-ion battery manufacture and disposal, mining and extraction, industrial catalysts, and pool/spa sanitizers. As battery demand grows, extraction and end-of-life disposal are worth watching as potential local contributors, though at present the natural background dominates.

From diet (a normal contributor)

Food also supplies trace lithium — grains and leafy greens among the richer sources. There is an active, unsettled scientific debate over whether trace dietary lithium is a beneficial micronutrient. That question is unresolved, and this page does not take a side; it is flagged here only so the water-route amount is seen in the context of everyday intake.

What the research says

The reason lithium in water attracts attention at all is a small body of population-level research suggesting trace lithium correlates with mental-health outcomes. According to PubMed, two lines stand out — and both come with hard limits on interpretation.

The suicide-rate association

Memon and colleagues (2020) in the British Journal of Psychiatry published a systematic review and meta-analysis of 15 ecological studies of naturally occurring lithium in drinking water and suicide rates. They found a consistent inverse (protective) association — pooled β = −0.27 (95% CI −0.47 to −0.08; P = 0.006) — meaning higher water lithium tracked with lower suicide rates across populations. Peer-reviewed The authors themselves caution that this is subject to the ecological fallacy: population-level correlations cannot establish that lithium caused anything in any individual. It is a signal that motivates research, not a basis for action.

The dementia association

Kessing and colleagues (2017) in JAMA Psychiatry ran a Danish nationwide case-control study (73,731 dementia cases; 733,653 controls). The relationship was nonlinear: compared with 2.0–5.0 µg/L, dementia incidence was lower at >15 µg/L (IRR 0.83) but higher at 5.1–10.0 µg/L (IRR 1.22). Peer-reviewed The non-monotonic shape and the study's own caveat — residual confounding cannot be excluded — are why this is intriguing rather than actionable.

What the toxicology shows

On the harm side, EPA's fact sheet is clear that documented adverse effects occur at therapeutic doses, not water levels: kidney effects (impaired urine concentration, and severe kidney disease with long-term higher-dose therapy), neurologic and endocrine effects (lethargy, tremor, cognitive impairment; thyroid and parathyroid effects), and rarer developmental effects. Regulatory EPA is refreshingly honest about the gap in between: it states it "cannot confidently estimate the risk" for water exposures between 10 µg/L and therapeutic-equivalent concentrations. The two-sidedness is real — a possible micronutrient benefit at trace levels versus clear harm at pharmacological levels — and the middle is genuinely uncharted.

What helps

If a well or system reports high lithium and you want to reduce it, reverse osmosis works. Per EPA, point-of-use RO reduces lithium by roughly 80–95%+, and ion exchange is effective at higher flow. This is the same toolkit that handles other dissolved metals in water.

Test rather than guess. Lithium is a cheap add-on to a standard water-quality panel. If you are on a private well in Western-US lithium geology, a one-time test tells you where you actually stand relative to the 10 µg/L screening level.

What does NOT help

Boiling. Per EPA, lithium is a dissolved metal — heating, boiling, or disinfecting does not remove it, and evaporation concentrates it.
Standard carbon pitcher filters. Activated carbon does little for a dissolved monovalent cation like Li⁺. Use RO or ion exchange if reduction is the goal.
Reading the UCMR 5 "most-detected" headline as a danger signal. High detection frequency reflects how widespread the mineral is, not that the levels are unsafe.
Equating water lithium with the drug. The therapeutic dose is ~1,000× higher; a Prop 65 listing for lithium carbonate is a drug-dose listing, not a water-level one.

Open questions

Whether trace lithium is a beneficial micronutrient at all — and if so, at what intake — remains genuinely unsettled. Speculation; the therapeutic-dose pharmacology is what is established
What happens biologically in the exposure window between 10 µg/L and drug-equivalent levels, where EPA says it "cannot confidently estimate the risk." Speculation
Whether pharmacological lithium's known circadian and sleep-architecture effects have any counterpart at trace water levels ~1,000× lower — there is no direct evidence that they do, and none should be implied. Speculation — mechanistic only; no sleep study supports a water-level effect
Whether growth in lithium-ion battery manufacture and disposal will meaningfully raise local groundwater levels over time. Inferred from the anthropogenic-source trajectory

Where you meet Lithium across your home

The same compound turns up in more than one place you live. Here's where it shows up in Embr — each links to the full breakdown for that part of your home.

Your tap water

Citations

US Environmental Protection Agency. Lithium in Drinking Water: A Resource for Primacy Agencies — HRL 10 µg/L; MRL 9 µg/L; no Health Advisory or standard; treatment. epa.gov Regulatory
California OEHHA, Proposition 65 List. Lithium carbonate — listed for developmental toxicity, 3/19/1999. oehha.ca.gov Regulatory
Memon A, Rogers I, Fitzsimmons SMDD, et al (2020). Association between naturally occurring lithium in drinking water and suicide rates: systematic review and meta-analysis of ecological studies. British Journal of Psychiatry, 217(6):667-678. DOI 10.1192/bjp.2020.128 Peer-reviewed
Kessing LV, Gerds TA, Knudsen NN, et al (2017). Association of Lithium in Drinking Water With the Incidence of Dementia. JAMA Psychiatry, 74(10):1005-1010. DOI 10.1001/jamapsychiatry.2017.2362 Peer-reviewed
Health Canada. Guidelines for Canadian Drinking Water Quality — Summary Table — no lithium guideline. canada.ca Regulatory
National Health and Medical Research Council. Australian Drinking Water Guidelines — no lithium value. nhmrc.gov.au Regulatory
US Environmental Protection Agency. Fifth Unregulated Contaminant Monitoring Rule (UCMR 5) occurrence data, 2023–2025 — 36.4% detection; median 23.3 µg/L; maximum 960 µg/L. epa.gov Regulatory

Frequently asked questions

Is lithium in drinking water dangerous?
At the levels found in US drinking water, there is no evidence it is harmful. Lithium is a naturally occurring mineral that dissolves into groundwater. Psychiatric lithium is dosed at 600 to 1,200 mg per day of a lithium salt; drinking water even at a high 100 µg/L delivers roughly 0.25 mg per day — about 1,000 to 4,000 times below a therapeutic dose. This is why the EPA set only a non-binding Health Reference Level of 10 µg/L rather than an enforceable limit. Documented harms — kidney and neurological effects — occur at pharmacological doses, not at water levels.
Why was lithium the most-detected substance in EPA's UCMR 5?
In EPA's Fifth Unregulated Contaminant Monitoring Rule (2023–2025), lithium was detected in 36.4% of 10,299 public water systems — more than any other analyte on the list — with a median detected concentration of 23.3 µg/L and a maximum of 960 µg/L. It is common because it is naturally abundant in geology and dissolves readily into groundwater, especially in arid Western-US aquifers over lithium-bearing formations. UCMR 5 is a screening program for unregulated contaminants, so a high detection frequency reflects how widespread the element is, not that it is unsafe.
Is the lithium in water the same as psychiatric lithium?
It is the same element, but the doses are on completely different scales. Prescription lithium (lithium carbonate) is given at 600 to 1,200 mg per day and produces measurable effects on mood, kidney function, and thyroid at those levels. Water-borne lithium exists as trace amounts of the Li+ cation. Across 1,464 US public-supply wells, EPA estimated a maximum daily ingestion of 0.05 to 1.0 mg from drinking 2.5 liters per day — versus a low therapeutic dose of 600 mg. The dose gap is the whole story.
How do I remove lithium from water?
Lithium cannot be removed by boiling, heating, or disinfecting the water. According to EPA, ion exchange is effective, and reverse osmosis reduces lithium by roughly 80 to 95 percent or more — so a point-of-use RO system is the straightforward choice if a well or system reports high values and you want to reduce them. A standard activated-carbon filter does little, because lithium is a dissolved monovalent cation that carbon does not adsorb well.

Related compounds

Embr researches the chemistry of where you live — including the water you drink. See the methodology page for how this Atlas tags claims by evidence strength, the tap-water hub for the other things in your water, and water filters compared for how to remove dissolved metals.

Last reviewed 2026-07-13. If you find a factual error, contact us.

Lithium in Drinking Water