How much lithium is in drinking water — and whether that is good or bad — depends entirely on context: local geology sets the amount, and the health signal shifts with dose and life stage. US groundwater is commonly grouped into bands of roughly ≤4, 4–10, 10–30, and >30 µg/L (USGS), with no federal limit. Population studies link higher levels to both lower suicide and, prenatally, higher autism risk, so there is no single verdict.
How much lithium is in US drinking water?
Lithium in US drinking water is measured in micrograms per liter (µg/L) and varies enormously by location and source. A USGS national assessment of public-supply and domestic wells found lithium ranging from below 1 to 396 µg/L, with a public-supply median of about 8.1 µg/L, highest in arid regions and older groundwater (Lindsey and colleagues, 2021; DOI). A later USGS-led model grouped estimated concentrations across the conterminous US into four bands, ≤4, >4–≤10, >10–≤30, and >30 µg/L (Lombard and colleagues, 2024; DOI).
The US Environmental Protection Agency is separately monitoring lithium in public water systems under its fifth Unregulated Contaminant Monitoring Rule (UCMR5). Results below the laboratory reporting floor appear as "non-detect." That floor is a method limit, not a health threshold; it is commonly cited near 9 µg/L.
Non-detect does not mean zero. A "non-detect" result means lithium was below the reporting floor, not that the water contains none. Many systems likely carry low single-digit µg/L levels that the method cannot quantify. Large cities served by surface water often fall into this low, non-detect range.
How does lithium in water vary by state and region?
Groundwater lithium tracks local geology: areas with certain rock and aquifer types carry more lithium, so concentrations vary by state, metro area, and even neighborhood within a region. Arid regions and certain aquifers tend to be higher, and surface water (rivers, reservoirs) often differs from the groundwater beneath the same area. Because the variation is geological rather than political, state-level averages hide large within-state differences, and the four occurrence bands above are a more useful frame than any single state figure.
How to read water lithium levels
The table below maps the USGS occurrence bands onto how those concentrations have appeared in research. It is an orientation aid, not a safety-threshold table: none of these bands represents an established safe or beneficial level.
| Band (USGS model, Lombard 2024) | How it appears in research context |
|---|---|
| ≤4 µg/L | Low; at or below roughly the 2 µg/L floor below which trace-lithium effects appear implausible in a dementia systematic review (Fraiha-Pegado and colleagues, 2024; DOI) |
| 4–10 µg/L | Low–intermediate; overlaps a band where one large Danish study reported higher dementia risk (5.1–10 µg/L; Kessing and colleagues, 2017; DOI) |
| 10–30 µg/L | Intermediate–high; includes levels associated with lower dementia risk in the same Danish study (above 15 µg/L; Kessing 2017) |
| >30 µg/L | High; still microgram-level, and roughly 1,000-fold below prescription lithium dosing |
Is lithium in drinking water good or bad?
There is no settled answer, and the honest position is that the question is genuinely unresolved. Population (ecological) studies have associated higher water lithium with lower suicide rates and, in some studies, lower dementia rates, but the dementia relationship is not linear: the largest Danish analysis found higher dementia risk in an intermediate band (5.1–10 µg/L) and lower risk only at the highest exposure (above 15 µg/L; Kessing and colleagues, 2017; DOI). At the other end, prenatal exposure carries a harm signal: a Danish study reported a dose-dependent association between maternal drinking-water lithium and autism risk in offspring (Liew, Ritz and colleagues, 2023; DOI).
None of these studies establishes causation, and there is no established beneficial or harmful threshold for ambient water lithium. The associations point in different directions at different concentrations and life stages, which is why no public-health body recommends either adding lithium to, or removing it from, household water for brain-health reasons.
The deeper evidence sits on the cluster pages: see lithium and dementia for the dementia epidemiology, lithium, mood and depression for the suicide-rate research, and the full studies database, by form to see how trace water-lithium evidence differs from carbonate and orotate evidence.
How do you test your water for lithium?
Lithium is not part of standard municipal water-quality reports, and reliable home test strips for it are not common, so testing generally means a certified laboratory. In brief: check whether your utility was monitored under UCMR5 and look up the USGS groundwater estimate for your area; determine whether you are on a public utility or a private well (private wells are unregulated and untested by default); send a sample to a laboratory accredited for trace-metal analysis; and interpret any result against the occurrence bands above, remembering that "non-detect" means below the method floor, not zero.
Full detail: how to test your water for lithium →
Do water filters remove lithium?
Whether a filter removes lithium depends on the technology. An EPA study of multiple water-treatment sites found that reverse osmosis removed more than 90% of lithium, while conventional surface-water treatment, adsorptive media, biological/aerobic groundwater treatment, and manganese-removal filters were not effective; cation exchange was inconsistent, and lime softening removed 11–54% (Keithley and colleagues, 2025; DOI).
| Filter or treatment type | Effect on lithium (Keithley 2025) |
|---|---|
| Activated carbon / conventional treatment | Not effective for dissolved lithium |
| Reverse osmosis (RO) | Removed more than 90% of lithium |
| Ion exchange (cation exchange) | Inconsistent removal |
| Lime softening | 11–54% removal (often still above 10 µg/L) |
That study assessed treatment-plant processes; the performance of consumer pitcher and faucet filters was not separately quantified, though carbon-based media did not reduce lithium. Common pitcher filters use activated carbon, so they are unlikely to lower lithium meaningfully.
Full detail: do Brita filters remove lithium? →
Limitations and safety
Water-lithium epidemiology is almost entirely ecological: it compares regional water concentrations with regional disease rates and cannot establish that water lithium causes any outcome in an individual. Reported associations are inconsistent, appearing protective, null, or (prenatally) harmful depending on the outcome and concentration, and regional confounders such as socioeconomic status, geology, and longevity are difficult to disentangle. There is no US federal maximum contaminant level for lithium in drinking water, and there is no evidence basis for deliberately adding lithium to, or removing it from, household water for brain-health reasons. This article is educational and is not medical advice.
For how trace water-lithium fits within the wider research on lithium and the brain, see lithium and the brain: the complete guide and dietary lithium intake.
Frequently asked questions
How much lithium is in tap water?
It varies widely. A USGS assessment found public-well lithium from below 1 to 396 µg/L, with a median near 8.1 µg/L (Lindsey and colleagues, 2021), and a USGS model groups estimates into ≤4, 4–10, 10–30, and >30 µg/L bands (Lombard and colleagues, 2024). Many public systems report "non-detect" under EPA UCMR5, meaning below the laboratory reporting floor (commonly cited near 9 µg/L), which is not the same as zero. Local geology is the main driver.
Is lithium in drinking water dangerous?
There is no established harmful threshold for ambient water lithium, and no US federal limit. Population studies link higher levels to lower suicide rates and sometimes lower dementia rates, but prenatal exposure carries an autism harm signal, and the dementia relationship is non-linear. None of these studies prove causation, so the net effect is genuinely unresolved.
Which states have the most lithium in water?
Lithium concentration follows geology, not state lines, and varies within states. Arid regions and certain aquifers tend to be higher. State averages hide large within-state differences, so the USGS occurrence bands described above are a more useful frame than any single state figure.
Does boiling water remove lithium?
No. Boiling does not remove dissolved lithium ions and can slightly concentrate them as water evaporates. Reverse osmosis is the most effective documented method (more than 90% removal), whereas conventional and carbon-based treatment are not effective (Keithley and colleagues, 2025).
Do Brita or carbon filters remove lithium?
Basic activated-carbon and conventional treatment were not effective at removing lithium in an EPA study, whereas reverse osmosis removed more than 90% (Keithley and colleagues, 2025). That study assessed treatment-plant processes rather than consumer pitcher filters specifically, but carbon-based media did not reduce lithium.
Should I add lithium to my water for brain health?
No. There is no evidence basis or safe established dose for adding lithium to household water, the population research does not prove benefit, and self-dosing lithium can be unsafe. Any interest in lithium should be discussed with a clinician.