The Claim
Two distinct health outcomes are attributed to radiofrequency (RF) transmissions from residential electricity smart meters:
- Cancer: ordinary RF exposure from installed smart meters increases cancer incidence among residents living nearby.
- Recurring headaches: smart-meter RF transmissions cause repeatedly occurring headaches in nearby residents.
The second outcome refers specifically to recurring headaches—not an undefined category of “smart-meter sickness”, insomnia, neurological disease or every chronic illness. A headache report after installation is an observation to investigate, not a measured cause.
Both components must be evaluated separately from the physical question of what RF fields the meter emits, which depends on design, duty cycle and location.
Where It Comes From
Smart-meter rollouts brought a new RF transmitter onto millions of buildings, sometimes without residents actively choosing the device. That combination of involuntary exposure, unfamiliar technology and utility-company communication created understandable concern.
Reports of symptoms after installation then became part of a broader electromagnetic-sensitivity and wireless-radiation debate.
Why It Sounds Convincing
The meter really does transmit RF energy, often from an exterior wall close to occupied rooms. A specification may list a maximum transmitter power that looks substantial when read without duty cycle or distance.
Consumer meters can also register brief peaks during transmission. A peak reading can be mistaken for the average exposure over hours or days.
The Short Answer
Smart-meter RF has not been established as a cause of cancer in nearby residents. The cited engineering and environmental experiments measure fields and transmission duty cycles; they are not long-latency cancer-incidence cohorts.
Smart-meter RF has not been established as a cause of recurring headaches either. Reports may be sincere and sometimes clinically important, but timing after installation does not demonstrate RF causation. The adjacent literature on short-term blinded EMF symptoms is relevant context, not a smart-meter-specific blinded headache trial.
Measured exposure is typically below applicable public RF reference levels in tested residential scenarios, yet duty cycle and measured fields differ significantly among smart-meter systems. None of those physical measurements can be converted directly into a disease-risk percentage.
How It Would Have to Work
Cancer endpoint: demonstrate measured or defensibly estimated meter-origin RF exposure, adequate latency, clinically ascertained cancer cases, and a reproducible dose-response relation after accounting for baseline risks and other exposures.
Recurring-headache endpoint: specify headache frequency, severity and observation window; compare on/off or different measured meter-RF conditions with blinding or suitably controlled observation where feasible. Control for knowledge of installation, expectation, noise, sleep, other environmental changes, medication and background headache frequency. A temporal association alone cannot isolate RF.
The physical emission measurements below establish how to characterize exposure, not either diagnosis.
What Actually Happens
Many smart meters send small packets of data periodically rather than broadcasting continuously at maximum rated power. The fraction of time spent transmitting—the duty cycle—can be low.
RF field strength also decreases with distance. Exterior placement and building materials further affect indoor exposure.
What the Evidence Shows
Smart meter is a device category, not one transmitter specification. Three original RF field studies illustrate why it is dangerous to assert that every smart meter has the same frequency, power, duty cycle or apartment exposure.
| Original measurement | System, real sample | RF result, correctly bounded |
|---|---|---|
| Addison, Calderon & Peyman (2023) | Three sampled UK metering system assemblies; 868-MHz communication hubs, in-home displays and gas-meter components under controlled laboratory conditions | Maximum 6-minute averaged 0.10 mW/m², duty factors under 2.8%. Authors compared with the ICNIRP 1998 reference for comparability with previous UK tests. The separate 2.4-GHz electricity meter did not form part of their 868-MHz subcomponent test. |
| Tell, Kavet & Mezei (2013) | One particular US utility’s 902–928-MHz frequency-hopping meter technology and operating logs from more than 88,000 meters | 99.9% of devices had duty cycle 1.12% or less. Peak readings at 0.3 m were at most 15% of the relevant US FCC public reference and samples inside six homes were at most 1%. 88,000 meters are not 88,000 human participants. Study supported by the Electric Power Research Institute (EPRI), with two authors affiliated with EPRI. |
| Aerts and colleagues (2020) | 915-MHz US utility single-meter through 81-meter-bank configurations, five urban field sites and laboratory facilities | Six-minute duty factors from 0.01% for a single meter to 13% for one 81-meter bank, a critical counterexample to claims that all installations transmit less than 1%. At 0.3 m, worst-case modeled/measured time-average levels in assessed installations were up to 9.43% of ICNIRP and 5.39% of FCC/IEEE reference. These numbers are not cancer-incidence percentages. |
A fourth original experiment, Foster and Tell (2013), separately studied 2.45-GHz Trilliant smart-meter architecture; its original abstract reports duty fractions below 1% under tested conditions but local indoor measurement was confounded by competing Wi-Fi/microwave RF signals. It does not independently validate the distinct 868/915-MHz configurations.
The main limitation across these experiments is decisive: the endpoints are power density, electric fields, peak readings, transmission fraction and regulatory comparisons. None tracked diagnosed cancers or chronic illnesses in residents, and ICNIRP 1998, ICNIRP 2020 and FCC have different applicable definitions. A typical low time-average result is useful exposure context; a particular elevated calibrated value warrants installation-specific assessment, not an automatic disease diagnosis.
Component A — cancer incidence
The key cited UK/US investigations do not contain cancer-incidence samples or long-latency participant follow-up. Their measurements constrain plausible exposure scenarios but cannot by themselves rule out every rare or delayed cancer outcome. The best current assessment is lack of demonstrated smart-meter RF cancer causation; broader RF oncology is indirect for installed meters.
Component B — recurring headaches
People may report headaches around the time of an installation. That experience merits appropriate health assessment, but neither temporality nor a meter peak establishes a causal trigger. Acute sham-controlled electromagnetic-sensitivity experiments can inform how expectation and actual exposure might be separated; they are not direct trials of smart-meter RF-induced recurrent headaches. The physical meter studies summarized above measured no participant headache endpoint.
A claim of recurring RF-triggered headaches would need a defined clinical/diary endpoint and a credible comparison that isolates RF from other changes.
The Strongest Evidence for the Claim
The strongest credible concern includes actual high-field configurations and uncertain long-term direct health outcomes. A 2020 primary physical field experiment found 13% duty cycles in an 81-meter bank, much higher than simplistic “all meters <1%” claims, and recorded distinct peaks near dense installations. Real symptoms following installation deserve clinical attention. This is meaningful counter-evidence to overgeneralized low-emission narratives, while not establishing that the RF caused cancer or those symptoms. The long-latency literature focused specifically on meter-to-person absorbed dose remains much thinner than the physical measurement base.
People can experience real symptoms temporally associated with installation, which should be taken seriously even when the causal trigger is uncertain.
The Strongest Evidence Against It
The strongest evidence is dosimetry. If a device is alleged to create a major radiation hazard, the actual field and absorbed exposure are the first quantities to test. Typical measurements are much lower than the claim implies.
The intermittent nature of transmission also undermines calculations that multiply maximum transmitter power by 24 hours as if the meter were continuously broadcasting at that level.
Common Viral Arguments
“The meter transmits all day.”
The device may be active all day, but RF transmission is typically intermittent. Active electronics and continuous maximum RF transmission are not the same thing.
“My meter gives a huge peak reading.”
A peak can be real. Health standards and cumulative exposure assessment also require units, distance, frequency and averaging time.
“Symptoms began right after installation.”
The timing deserves attention but cannot identify the cause by itself. Blinded and comparative evidence is needed to separate RF from expectation or other environmental changes.
“There are no decades-long smart-meter trials.”
That is a limitation of device-specific evidence. It does not convert low measured exposure into positive evidence of disease.
The Reasoning Error
Hazard-risk confusion treats any measurable RF emission as evidence of harmful exposure. Dose neglect omits duty cycle, distance and actual field strength.
Correlation-causation error can convert symptom timing after installation into a demonstrated RF mechanism without a control comparison.
What Would Change Our Conclusion?
For cancer, replicated, latency-appropriate studies linking well-characterized residential meter RF to clinically verified cancer outcomes, with credible controls, could change the conclusion.
For recurring headaches, blinded or otherwise well-controlled, independently reproduced studies that link a specified headache endpoint to measured meter-origin RF—not simply knowledge of meter installation—could change the outcome-specific assessment. New meter models with substantially different emissions would first require exposure measurement; an increased peak reading alone would not establish clinical harm.
Evidence Status
Status: Unsupported
Evidence quality: Moderate
The cited studies document RF emissions but do not demonstrate either claimed clinical cause.
- Cancer component — Unsupported: no smart-meter-specific causal increase in residents’ verified cancer incidence is established; the exposure literature is more direct than the required long-term clinical evidence.
- Recurring-headache component — Unsupported on current direct evidence: installation-timed symptom reports and nearby EHS provocation research do not constitute meter-specific headache causation. The direct clinical evidence base for this particular outcome is limited, so this is not proof that every unusual response is impossible.
The Moderate evidence-quality grade describes the available exposure measurements and indirect risk context; direct outcome evidence is substantially weaker and should not inherit that label unqualified. The 88,000-plus US telemetry records counted meters, not people.
Sources
This investigation emphasizes direct exposure measurement and technical assessment while keeping the narrower limitation—less extensive smart-meter-specific long-term epidemiology—visible.
Corrections & Updates
October 9, 2026 — evidence and scope clarification. Separated the cancer-incidence question from recurring headaches while retaining the originally examined radiofrequency (RF) measurement contrasts, Electric Power Research Institute (EPRI) sponsorship and the 81-meter-bank 13% duty-cycle counterexample. The >88,000 logs were from devices, not patients; International Commission on Non-Ionizing Radiation Protection (ICNIRP) 1998/2020 and US Federal Communications Commission (FCC) quantities are not interchangeable. The existing Unsupported / Moderate umbrella judgment remains unchanged.
Last Reviewed
October 9, 2026.