Yes, smart meters emit radiofrequency (RF) radiation, but at power densities far below the levels that produce biological heating, and typically for only seconds to minutes per day. The main chipset transmits in brief bursts at roughly 1 watt or less. That puts everyday exposure from a meter many times lower than from a mobile phone call. The open question is not whether the meters radiate; they do. It is whether the pattern and duration of that radiation matter.
The Transmission Mechanism
A smart meter is a small radio transceiver bolted where the gas or electric line enters a building. Inside, a radio module converts digital readings into a modulated RF signal and pushes it through an antenna. The antenna is usually a small patch or wire element inside the housing. Most North American utility meters operate in licensed or unlicensed bands near 900 MHz. Some use the 2.4 GHz band shared with Wi-Fi. The exact frequency depends on the utility's chosen network, not on the meter's brand alone.
The mechanism that matters biologically is dielectric heating. An RF field exerts torque on polar molecules, chiefly water, and that molecular motion converts to heat. The depth of penetration shrinks as frequency rises, so 900 MHz and 2.4 GHz fields deposit energy mostly in the first few centimeters of skin and underlying tissue. The body's thermal regulation usually dissipates that heat without measurable temperature change. Biological effects that appear without heating remain unconfirmed and are not the basis for any public exposure limit.
Smart meters also differ from older analog meters in one crucial way. An analog meter is passive: it spins a disk and radiates nothing. A smart meter is an active transmitter that reports consumption at intervals. That is the source of every credible exposure concern, and it is why duration and duty cycle matter more than raw power.
Quantifying the Exposure
Standardized measurements of smart meters find that average power density falls well below 1 milliwatt per square centimeter (mW/cm²) at typical distances. The FCC general-population limit for the same bands is also about 1 mW/cm², averaged over 30 minutes. The asymmetry is the headline. A meter operating at its peak sits at or below the legal ceiling for continuous public exposure and falls orders of magnitude below it on a time-averaged basis.
Here is the mechanism in practice. A meter typically transmits in short bursts, often under a second, and many models are scheduled to report only every several minutes to a few times per day. Averaged over 24 hours, the radio is effectively idle for the vast majority of that window. That duty cycle is what separates a meter from a continuously connected device such as a Wi-Fi router or a cell tower.
Distance obeys the inverse-square law. Double the separation from the antenna and the power density drops to roughly one quarter. At one meter, the field from a typical meter is already far lower than at ten centimeters. A wall between the meter and occupied space adds attenuation that further cuts the field, more at 2.4 GHz than at 900 MHz because higher frequencies lose energy faster through masonry and drywall.
For calibration, a mobile phone held to the head can exceed the same regulatory limits in absolute transmitted power, though phones dynamically reduce power when signal is strong. A smart meter is a lower-power, intermittent device. A microwave oven leaks a regulated maximum of 5 mW/cm² at 5 cm at manufacture, and real-world use stays far below that.
Timeline of a Meter's Daily Radio Activity
The time course is useful because duty cycle, not peak power, dominates the real exposure. The rows below describe typical behavior across meter classes, not a single installed unit.
| Period | Meter state | Radio behavior | What a resident sees |
|---|---|---|---|
| Boot and network join | Power-up | Brief higher-power sync and handshake | No visible sign |
| Routine reporting | Steady state | Short bursts at scheduled intervals, often minutes to hours apart | Nothing |
| Demand or event response | On-demand | Occasional extra bursts triggered by outage detection or pricing events | Nothing |
| Firmware update | Maintenance | Longer, infrequent download session | Nothing |
| Peak load periods | Steady or event | Unchanged for most residential meters | Nothing |
| Night rest window | Idle | Some utilities restrict transmissions | Nothing |
The honest caveat: reporting intervals are a utility policy decision, not a fixed property of smart meters. A meter programmed to report every 15 seconds is a different exposure profile than one reporting every 15 minutes, even with identical hardware. Ask the utility for its duty-cycle specification before assuming a number.
Edge Cases and Higher-Exposure Scenarios
Two installations concentrate the field more than a wall-mounted residential meter. First, a bank of meters mounted together in a corridor or utility closet creates overlapping sources. The field at the face of such an array is higher than from any single unit, and standing directly in front of it for long periods concentrates exposure. Second, a meter on a metal backplate near a seating area at close range is closer to a resident than the assumed one-meter separation. Neither scenario typically breaches the public limit, but both push the margin down and are worth knowing.
Water and gas meters in some systems use lower-power radios or walk-by reading, while electric meters on mesh networks may relay neighbors' traffic as well as their own. Relaying raises the duty cycle. A meter acting as a network node transmits more often than a leaf node, so a utility that deploys a mesh architecture can have some meters doing more radio work than others. The user has no visibility into which role a given meter plays.
Meters on the exterior of a house radiate away from occupied space. Meters inside a garage or utility room radiate into it. Placement, not just power, governs what reaches a person.
Misconceptions and What the Evidence Shows
One common claim holds that smart meters cause symptoms such as headaches, sleep disruption, or fatigue. The mechanism such claims rely on is unconfirmed, and blinded provocation tests conducted in the early 2010s with self-identified electrosensitive participants generally failed to separate real fields from sham exposure. People did not reliably detect whether the meter was on. That outcome does not prove everyone is insensitive, but it undercuts the claim that a specific, detectable signal drives the symptoms.
A second misconception treats "radiation" as a single hazard word. RF is non-ionizing. It lacks the photon energy to break chemical bonds or damage DNA directly the way X-rays or gamma rays can. The only established mechanism at ordinary levels is heating. The World Health Organization and the FDA review the evidence and set exposure guidance on that basis. Conflating a smart meter with a medical X-ray is a category error, not a risk estimate.
A third claim confuses peak with average. Critics often cite a meter's transmitted power and compare it to a limit that is already averaged. The number that matters for biological effect is the time-averaged field where a person sits, and that is lower than the peak by the duty cycle.
Practical Actions
Residents who want to reduce exposure to near zero have low-cost options, and none require special equipment beyond what the utility provides.
Request the utility's meter model and its published power and duty-cycle specification, then calculate the average field at your seating distance using the inverse-square relationship. Ask whether the utility offers an opt-out or a non-transmitting electromechanical meter, and confirm any fee in writing. Relocate a meter installed indoors or in a utility closet to the exterior wall if the utility allows it. Keep beds and long-duration seating at least one meter from a meter bank or indoor meter face; every doubling of distance cuts the field to about a quarter. Red-flag: refuse to open or remove a meter yourself, because tampering is both a safety and a legal hazard. Red-flag: treat an opt-out as full protection only for that meter's radio; a mesh network may still transmit from a neighbor's unit.
Reconciling the thresholds people encounter: the FCC public limit near 1 mW/cm² is an averaged regulatory ceiling, the much stricter limits some other countries set apply their own averaging and distance assumptions, and no official guidance sets a separate "smart meter" figure at all. Makers therefore certify against the general public rule. The honest position is that meter exposure is a small fraction of a limit already designed with a wide safety margin.
The counterintuitive detail: for a resident standing two meters away, a smart meter's time-averaged exposure often constitutes a smaller share of the day's RF dose than the Wi-Fi router in the kitchen or the phone in a pocket, despite the meter's reputation as a focused emitter. The reputation comes from visibility, not from measured power.
Limitations of What Can Be Concluded
The evidence base has real limits, and they cut in both directions. Meter-specific dosimetry studies exist but are few, and they model individual units rather than the many network architectures utilities deploy. Duty cycles are configurable and often undisclosed, so any published average can be wrong for a specific home. The biological literature on non-thermal RF effects is large and noisy, and the absence of a confirmed mechanism does not equal proof that none exists; it means no effect has been established at these levels. For anyone with an implanted medical device, the relevant question is device immunity rather than general population risk, and that should be answered by the device manufacturer's guidelines rather than by a general meter measurement.