The Rodenticides Nobody Measures in Water

What we went looking for

Our analysis of anticoagulant rodenticides in California wildlife reported that California’s state laboratory found residues in 128 of 158 wild animals tested in 2022, with brodifacoum, bromadiolone, difethialone and diphacinone among the most common. Our analysis of insecticides in US streams drew on tens of thousands of water samples per compound.

The obvious next question was whether the same rodenticides appear in US surface water. We expected to report a detection rate. Instead we found that, for most of them, the national record cannot hold a measurement at all.

Six compounds that are not in the database at all

The Water Quality Portal stores every result under a registered characteristic name. A compound without a registered name cannot have a result stored against it, regardless of whether anyone has measured it. We queried the portal’s own name service for each rodenticide used in structural pest control, and for three insecticides as a reference.

Source: Water Quality Portal characteristic name service and Result search, queried 2 October 2026. Author compilation.

What it means. Brodifacoum, the compound California’s wildlife laboratory most often names among the residues it finds, has no registered name in the national water-quality database. Neither do difethialone, difenacoum, diphacinone, chlorophacinone or bromethalin. For comparison, fipronil has nine registered names, because the database also tracks its breakdown products separately.

What limits it. An unregistered name means no result has been submitted to this database under that compound. It does not prove no laboratory anywhere has ever measured it. Academic studies that never submit to the portal would not appear. The correct statement is that the national public record holds nothing for these six.

Warfarin is measured, but only as a medicine

Warfarin has 9,710 results in the national record, 4,345 of them from streams. At first sight that looks like the exception. It is not, for a reason worth stating precisely.

Warfarin is both a first generation rodenticide and a widely prescribed blood thinner. Every method used to measure it in US streams is a pharmaceutical method. The three most used are a pharmaceutical liquid chromatography method applied to 1,811 results, a method for pharmaceuticals and personal care products in water, EPA Method 1694, applied to 1,432 results, and a second pharmaceutical schedule applied to 614. None is a pesticide method.

What it means. Warfarin in US streams is monitored as a wastewater pharmaceutical. Those results cannot be attributed to rodenticide use, because the monitoring was not designed to distinguish the two sources and a medicine taken by millions of people is a far larger plausible input. A second registered name, Warfarin-d5, is a deuterated laboratory standard rather than an environmental measurement.

What limits it. We did not attempt to apportion warfarin between medical and rodenticide sources, which these data cannot support.

Bromadiolone: one year, two programmes, four uncertain values

Bromadiolone is the only anticoagulant rodenticide with a pesticide-method record, and it is very small.

Source: Water Quality Portal Result search for Bromadiolone, all media, retrieved 2 October 2026.

What it means. A naive count would report 11 values. Seven are quality-control spike recoveries, which measure how well the laboratory recovered a known amount of compound it added deliberately. They are not environmental measurements. Counting them as detections would be a serious error.

The four remaining values are reported as concentrations of 59.9, 77.0, 78.5 and 78.7 nanograms per litre, all from one programme in 2009, and all flagged as estimated rather than actual. Those values fall in the same range as the same programme’s recovery percentages, which run from 78.5 to 85.0 in the main cluster. We cannot determine from the record whether these four are genuine environmental detections or recovery values entered with the wrong unit, so we report them as unconfirmed.

What limits it. We are not asserting a data entry error. We are stating that the record does not allow these four values to be confirmed, which is the relevant fact for anyone tempted to cite them.

Why this matters alongside the wildlife evidence

The contrast is the finding. California’s state laboratory detected anticoagulant rodenticide residues in 81.0 percent of the 158 wild animals it tested in 2022. The national water record holds no confirmable pesticide-method detection of any of the compounds involved.

Those two facts are not contradictory. Anticoagulant rodenticides reach wildlife primarily through the food chain, as predators eat poisoned rodents, rather than through water. A compound can be widespread in wildlife and absent from water monitoring because the water pathway was never the main route. But the absence also means the water pathway has been assumed rather than measured. Nobody can currently cite national data showing these compounds are not in surface water, because nobody has submitted the measurements that would show it.

What limits it. The food-chain explanation is the standard account of how these compounds reach wildlife. We present it as context, not as something these data demonstrate.

What absence does and does not mean

Three checks we applied

1. Registration check before reporting a zero

Definition. Confirming that a compound has a registered name in the database before interpreting an empty query result.

Method. Query the portal’s characteristic name service for the compound. If no name is registered, an empty result means the database cannot hold the measurement, not that the compound was not found.

Result. Six of eight rodenticides failed this check. Reporting their empty results as “not detected” would have been wrong.

Limitations. Name matching is by text and could miss an unusual synonym. We searched the common name of each compound.

2. Method attribution

Definition. Checking which analytical method produced a compound’s results, to establish what the monitoring was designed to detect.

Method. Group results by the analytical method name field.

Result. Every warfarin stream result came from a pharmaceutical or personal-care-product method. None came from a pesticide method.

Limitations. Method name indicates design intent, not the actual source of the compound in the water.

3. Quality-control separation

Definition. Separating laboratory quality-control records from environmental results before counting detections.

Method. Exclude results whose unit is percent recovery, then examine the value type of what remains.

Result. Seven of eleven valued bromadiolone results were spike recoveries. The four remaining were all estimated.

Limitations. A recovery value entered with a concentration unit would pass this filter, which is precisely the ambiguity we report.

Related: how rodenticide use in California pest control shifted after the 2021 restriction is in California’s rodenticide ban, measured.

Related research: insecticides in US groundwater.

Download the data

The tables in this report are published as CSV files under a Creative Commons Attribution 4.0 licence. You may republish, remix and build on them, including commercially, with attribution to The Pest Control Brief and a link to this page.

• Registration status, result counts and methods for eight rodenticides and three reference insecticides (CSV, 11 rows)
• Every valued bromadiolone result in the national record, classified (CSV, 11 rows)

How to cite this report

The Pest Control Brief, “The rodenticides nobody measures in water”, 2 October 2026. https://thepestcontrolbrief.com/rodenticides-not-measured-in-water/

References

1. Water Quality Portal, characteristic name service. waterqualitydata.us/Codes/characteristicname, brodifacoum. Queried 2 October 2026 for each compound named in this report.
2. Water Quality Portal, Result search for Warfarin, stream sites. waterqualitydata.us Result search, Warfarin. Retrieved 2 October 2026.
3. Water Quality Portal, Result search for Bromadiolone, all media. waterqualitydata.us Result search, Bromadiolone. Retrieved 2 October 2026.
4. US Environmental Protection Agency. Method 1694: Pharmaceuticals and Personal Care Products in Water, Soil, Sediment and Biosolids by HPLC/MS/MS. The method under which 1,432 warfarin stream results were produced.
5. California Department of Fish and Wildlife, Wildlife Health Laboratory. 2022 Summary of Pesticide Exposures and Mortalities in Non-Target Wildlife, published 31 July 2023. nrm.dfg.ca.gov document 241290. Source of the wildlife exposure figures.

How this was established

Approach. For each of eight rodenticides used in structural pest control, we queried the Water Quality Portal’s characteristic name service to establish whether a result could be stored. For the two with registered names, we downloaded every result and examined the contributing programmes, analytical methods, years, units, detection conditions and value types.

Inclusion. All media for bromadiolone, because the record is too small to subset. Stream sites for warfarin, to match our other water analyses.

Quality-control handling. Results reported in percent recovery were classified as laboratory spike recoveries and excluded from any detection count. Deuterated standards such as Warfarin-d5 were classified as laboratory standards.

What we did not do. We did not search academic literature for studies that never submitted to the portal, so this report describes the national public record, not all research. We did not apportion warfarin between medical and rodenticide sources. We did not attempt to resolve the four unconfirmed bromadiolone values.

Limitations. The Water Quality Portal holds what contributing programmes submit, so absence reflects monitoring choices as well as environmental reality. Name matching is by common name. The food-chain explanation for wildlife exposure is presented as established context rather than demonstrated here.

Last reviewed: 2 October 2026.

Sources consulted

Tier 1, primary government data. Water Quality Portal, operated by the US Geological Survey, the US Environmental Protection Agency and the National Water Quality Monitoring Council. California Department of Fish and Wildlife, Wildlife Health Laboratory.

Excluded. Laboratory quality-control recoveries and deuterated standards, which are not environmental measurements.

For reporters covering rodenticides

Most quotable statistics

1. Six of eight rodenticides used in pest control, including brodifacoum, have no registered name in the US national water-quality database.
2. All 4,345 warfarin results from US streams were produced by pharmaceutical methods, not pesticide methods.
3. The entire national record for bromadiolone is 56 results, all from 2009.
4. Of 11 bromadiolone results carrying a value, 7 were laboratory quality-control recoveries.
5. California’s wildlife laboratory found anticoagulant rodenticide residues in 81.0 percent of 158 animals tested in 2022, while the national water record holds no confirmable detection of the compounds involved.

Figures to quote alongside them

Absence from the national record means no measurement was submitted, not that the compounds are absent from water. Anticoagulant rodenticides reach wildlife mainly through the food chain. Warfarin is also a prescription drug.

Fields in the published CSV files

Compound; class; registered name count; registered names; national results across all media; results from streams; measurement methods; note. For bromadiolone: row, organisation, value, unit, value type and classification.

Press summary

The rodenticides most often found in wildlife are almost entirely absent from the US national water-quality record, according to an analysis by The Pest Control Brief. Six of eight rodenticides used in structural pest control, including brodifacoum, have no registered name in the Water Quality Portal, which means no measurement of them can be stored. Warfarin, the one widely measured, is monitored only as a pharmaceutical, with every stream result produced by a pharmaceutical method. Bromadiolone’s entire national record is 56 results from 2009, and of the 11 carrying a value, 7 are laboratory quality-control recoveries. California’s wildlife laboratory found anticoagulant rodenticide residues in 81 percent of animals it tested in 2022. The analysis concludes that the water pathway for these compounds has been assumed rather than measured.

Five headlines editors could use

1. The rodenticides nobody measures in water
2. Found in 81 percent of tested wildlife, absent from the national water record
3. Brodifacoum has no name in America’s water-quality database
4. Warfarin is in US streams, but it is being measured as a medicine
5. Why a zero in pesticide data can mean nobody looked

What readers ask about this

Are rodenticides in US rivers and streams?

The national record cannot answer that for most of them. Six of eight have no registered name in the database, so no measurement can be stored. Bromadiolone was measured in one year, 2009, without a confirmable detection.

Does this mean rodenticides are not polluting water?

No. It means the water pathway has not been measured nationally. Absence of measurement is not evidence of absence.

Why is warfarin measured when the others are not?

Because warfarin is also a widely prescribed blood thinner and is monitored as a pharmaceutical in wastewater-affected streams. Every warfarin stream result came from a pharmaceutical method.

How do rodenticides reach wildlife if not through water?

Mainly through the food chain, when predators and scavengers eat poisoned rodents. That is the standard account, presented here as context.

Were the 2009 bromadiolone values detections?

Not confirmably. Four values were reported as concentrations, all estimated, and they fall in the same range as the same programme’s quality-control recovery percentages.

What is a quality-control spike recovery?

A laboratory check in which a known amount of a compound is added to a sample to measure how much the method recovers. The result is a percentage, not an environmental concentration.

Could researchers have measured these compounds outside this database?

Yes. This report covers the national public record. Academic studies that never submitted results would not appear.

What does this mean for pest control operators?

That the evidence base on rodenticides in the environment rests heavily on wildlife testing rather than water monitoring. Our coverage of rodenticides in California wildlife and restricted use pesticides sets out the regulatory context.

Which compounds are measured well in water?

Several insecticides are. See our audit of national monitoring data for which detection rates are robust.

Can I check this myself?

Yes. Query the Water Quality Portal characteristic name service for any compound. If it returns no registered name, the database cannot hold a result for it.