01 · The Limits
The safety limits are not a guess. They already contain a large safety margin.
RF exposure limits in the United States and internationally aren't a single agency's opinion — they're set independently by multiple bodies who arrive at the same order of magnitude, using thermal and behavioral thresholds established in controlled research, and then built in a substantial safety factor on top:
FCC (United States)
Adopts limits based on IEEE/ICNIRP guidance. For general public exposure above 6 GHz, the limit is 1 mW/cm² averaged over 4 cm² and 30 minutes; localized exposure (e.g. near the head) is capped at 2 mW/cm² over 1 cm². Below 6 GHz, limits are expressed as Specific Absorption Rate (SAR), capped at 1.6 W/kg averaged over 1 gram of tissue for the general public.
FCC: Guidelines for Cellular Antenna Sites →ICNIRP (International)
The International Commission on Non-Ionizing Radiation Protection published updated guidelines in March 2020 after a multi-year review of the scientific literature. General public reference levels above 6 GHz sit at 10 W/m² — one-fifth the occupational limit of 50 W/m². Sub-6 GHz limits are essentially unchanged from the 1998 guidelines, reflecting decades of consistent evidence at those frequencies.
ICNIRP: 2020 RF Guidelines →Health Canada & ARPANSA
Canada's Safety Code 6 and Australia's ARPANSA standard are developed independently of the FCC and ICNIRP, by separate national regulators with their own review processes — and land on comparable exposure limits, which is exactly what you'd expect if the underlying biophysics, not politics, is driving the number.
Health Canada: Safety Code 6 →Why the margin matters: these limits are set well below the threshold where any established biological effect (tissue heating) begins — commonly cited as roughly a 50x safety factor for the general public relative to the level where measurable heating starts. A device that reads "close to the limit" is not close to a level with any demonstrated health effect.
02 · Real-World Exposure
Actual ambient RF levels are a tiny fraction of the limit — that's in the FCC's own field data.
The FCC has stated plainly, based on field measurements near typical cellular and PCS installations:
"Measurements made near typical cellular and PCS installations, especially those with tower-mounted antennas, have shown that ground-level power densities are hundreds to thousands of times less than the FCC's limits for safe exposure... it is extremely unlikely that a member of the general public could be exposed to RF levels in excess of FCC guidelines due solely to cellular or PCS base station antennas located on towers or monopoles." — FCC, Human Exposure to Radio Frequency Fields: Guidelines for Cellular Antenna Sites
This is the context missing from most viral videos of someone waving a meter near a router: the number on the screen being "high" relative to a nearby harmless source tells you nothing about where it sits relative to the actual safety limit, which is typically hundreds to thousands of times higher than what's measured in ordinary environments.
Read the full FCC guidance →03 · The "10 Microwatt" Claim
The "10 microwatt" limit isn't a real, unified safety threshold — it's several unrelated numbers wearing the same units
A specific figure — usually stated as "10 microwatts per square centimeter" or sometimes just "10 microwatts" — circulates as if it were an internationally recognized safe limit that the FCC is ignoring. It doesn't hold up once you check where each number actually comes from and what unit it's actually in.
The units get quietly swapped
The Salzburg guideline — one of the most-cited "precautionary" figures — sets a target of 10 microwatts per square meter (µW/m²) outdoors, not per square centimeter. That distinction matters enormously: one square meter is 10,000 square centimeters, so 10 µW/m² converts to 0.001 µW/cm² — a number four to five orders of magnitude smaller than what gets repeated online as "10 microwatts." Fringe content routinely drops or swaps the unit, which turns an already-conservative advocacy figure into something that sounds like an official international consensus limit, when it's neither official nor was it ever the figure being cited.
These are precautionary advocacy figures, not derived safety thresholds
The Salzburg guideline and the BioInitiative Report's recommended cautionary levels were not derived from a demonstrated harm threshold the way FCC/ICNIRP limits were. They apply the precautionary principle — recommending exposure be minimized well below any established effect, as a hedge against effects that have not been established. That's a defensible policy stance to advocate for, but it's a different claim than "10 µW/cm² is where harm starts and the FCC limit is unsafe." The BioInitiative Report itself was formally reviewed by the Health Council of the Netherlands in 2008, which concluded it was a selective review that did not present a balanced assessment of the relative scientific quality of the studies it drew on.
A few countries do have stricter enforceable limits — and even those are far above the viral figure
Italy and Russia have adopted actual regulatory limits around 10 µW/cm² for some exposure categories — genuinely stricter than the FCC's roughly 580–1000 µW/cm² general-public limit for cellular frequencies, by a factor of roughly 60–100x. That's a real, citable policy difference between countries, reflecting different regulatory philosophies. But it's still 10,000x higher than the 0.001 µW/cm² figure the Salzburg number actually works out to, and no version of these stricter national limits was set because controlled research demonstrated a health effect at FCC/ICNIRP levels — it reflects a different tolerance for uncertainty, not new evidence of harm.
The takeaway: when you see "10 microwatts" cited as the "real" safe limit, ask which number it actually is, and in which unit. The versions that are genuinely precautionary (Salzburg, BioInitiative) are separated from the FCC limit by many orders of magnitude and were never derived from a demonstrated harm threshold in the first place — they're a policy choice to be more cautious than the evidence currently requires, which is a legitimate position, but not the same claim as "current limits are proven unsafe."
04 · The Rat Studies
The NTP and Ramazzini studies, without the misleading summary
These two studies are the most scientifically serious evidence RF-safety skeptics cite, and they deserve a straight answer rather than dismissal. Here's what they actually did and what health agencies concluded from them.
What the studies did
The U.S. National Toxicology Program (NTP) and Italy's Ramazzini Institute each exposed thousands of rats and mice to RF radiation for most or all of their lives — far longer and more continuous exposure than any human phone user experiences — and found a statistically elevated rate of certain tumors (malignant gliomas, heart schwannomas) in male rats at the highest dose groups.
Why the dose comparison is the crux of the debate
Critics of the studies note the tested Specific Absorption Rate (SAR) levels reached or approached the localized human exposure limit (1.6 W/kg) — not ambient background levels, which are far lower. Defenders of the studies counter that localized tissue near a phone antenna can, in fact, receive exposure in that range during a call. The disagreement is real and unresolved among researchers — but note what it is not a disagreement about: nobody in this exchange claims ambient RF from towers, WiFi, or a phone in your pocket is remotely close to these dose levels.
How major health bodies read the evidence, as of the most recent reviews
The National Cancer Institute and American Cancer Society both describe the NTP/Ramazzini findings as notable but not sufficient, on their own, to establish that typical RF exposure causes cancer in humans — citing the very high and prolonged exposure levels used, the lack of a dose-response relationship in several endpoints, and the absence of a corresponding rise in glioma or related cancer rates in the human population despite decades of mass cell phone adoption. The IARC's 2011 classification of RF as "Group 2B — possibly carcinogenic" (the same category as aloe vera extract and pickled vegetables) reflects limited, not established, evidence and predates both the NTP and Ramazzini final reports.
05 · Consumer RF Meters
What a "Safe & Sound Pro II" reading actually tells you: not much
The technical problem
Consumer and "prosumer" broadband RF meters — the Safe & Sound Pro II/Classic, Cornet, Acoustimeter, TriField, and similar devices — are not laboratory spectrum analyzers, and they have real, documented limitations that matter for the conclusions people draw from them:
- They're broadband, not frequency-selective. A single antenna and detector sum up energy across a wide frequency range at once. That means the device can't distinguish a momentary WiFi handshake, a microwave oven door seal leak, a Bluetooth beacon, and continuous cell signal — they all just add to one number.
- Published accuracy is wide even for the "better" models. The Safe & Sound Pro II's own manufacturer-cited third-party lab testing states accuracy of roughly ±6 dB. That sounds precise; in linear power terms it's not — a ±6 dB uncertainty band corresponds to roughly a 4x range between the low and high end of what the true power density could actually be for a given reading. Independent RF engineering references note that getting within 3 dB (a 2x range) in real-world handheld RF measurement is already considered a good result, because reflections, body proximity, and antenna orientation all distort the reading.
- Frequency weighting varies device to device. Because no two meters share identical antenna and detector characteristics, two different "EMF meters" pointed at the same source at the same moment routinely disagree — sometimes significantly — with neither reading being wrong given how each device is built, just built differently.
- No legally meaningful averaging window. Regulatory limits are defined as an average over specific time windows (30 minutes for general public exposure, per FCC/ICNIRP) and specific areas (1–4 cm²) — precisely because RF fields fluctuate constantly and instantaneous peaks are not what determines biological relevance. A handheld meter showing a needle jump for half a second is showing you something the regulatory limit was never designed to be judged against.
- Frequency range gaps. Many popular consumer models top out around 6–8 GHz, meaning higher 5G millimeter-wave bands aren't measured by the device at all — so a "safe" reading may simply reflect a blind spot in what the device can see, not the absence of RF energy.
The business model problem
A meter reading is rarely sold on its own. It's frequently the first step of a sales funnel that ends with a "board certified EMF consultant" recommending mitigation products — shielding paint, grounding mats, "quantum" stickers, phone radiation patches — priced from tens to thousands of dollars. The Federal Trade Commission has repeatedly taken enforcement action against sellers of these mitigation products for making claims with no scientific support:
"The complaint alleged that the defendants failed to disclose that most of the energy emitted by cellular and cordless phones comes from the antenna and other parts of the phone, and that [the shielding product] has no effect on these other emissions... there is no scientific proof that so-called shields significantly reduce exposure." — FTC, re: Comstar Communications / WaveShield settlement
The pattern repeats across two decades of FTC actions: a device (often a consumer RF meter, sometimes just a claim) is used to create alarm, followed by a paid product or consultation to resolve the alarm the device itself created.
What a meter is actually good for: identifying the relative location and rough presence of an RF source (useful for, say, finding where a WiFi router is strongest in a room). What it cannot responsibly tell you, given the accuracy ranges above, is precisely how close you are to a regulatory safety limit, or whether a given reading has any health significance at all.
06 · Local Government Authority
What cities can and can't do about 5G — the actual federal law, not the version circulating online
A common claim is that local governments have "no say" over 5G deployment, or conversely that federal law is a secretive industry giveaway with no legal basis. Neither is accurate. The actual rule is narrower than either version, is a 30-year-old act of Congress, and has been repeatedly tested and upheld in federal court.
What Section 704 actually preempts — and what it doesn't
Section 704 of the Telecommunications Act of 1996, codified at 47 U.S.C. § 332(c)(7)(B)(iv), says a state or local government may not deny a wireless facility siting application "on the basis of the environmental effects of radio frequency emissions" if the facility complies with FCC exposure limits. That is a narrow, specific restriction: it removes RF-health justifications as a legal basis for denial. It does not touch — and local governments retain full authority over — zoning, placement, height and setback rules, structural and fire safety, historic-district and aesthetic review, and permit fees. A city can still deny or condition a cell site application on virtually every ground except "we're worried about the radiation."
This has been litigated, and it held up
The provision's constitutionality was challenged directly in Cellular Phone Taskforce v. FCC (2d Cir. 2000), which upheld it; the Supreme Court declined to hear a further appeal. Courts have consistently read Section 704 as leaving local zoning authority intact while foreclosing RF-emissions/health as the stated rationale for a denial — the balance Congress struck deliberately, not one federal regulators invented unilaterally later.
The 2018 "shot clock" order is a separate, later rule — and it lost part of a court challenge
In 2018 the FCC adopted rules setting time limits ("shot clocks," generally 60–90 days) for local review of small wireless facility applications and capped certain fees, aimed at preventing indefinite delay. Multiple cities, including Portland, sued. The Ninth Circuit's 2020 ruling in City of Portland v. FCC upheld most of the order — including the shot clocks — but struck down the portion that tried to dictate how localities must evaluate aesthetic conditions, finding that piece arbitrary and capricious. In other words: even the FCC's more aggressive attempt to speed up deployment was not a blank check, and a federal court trimmed it back where it overreached. That's the opposite of an unaccountable process.
The honest summary: local governments cannot legally reject a compliant wireless facility because they're worried about radiation — that specific veto was removed by Congress in 1996 and upheld in court. Local governments retain essentially every other form of zoning and permitting authority they had before, and federal attempts to further streamline review have themselves been subject to — and partially lost — court challenges. Claims that cities have "no power" or that the law is unconstitutional and unchallenged are both wrong.
07 · Sources
Primary sources cited on this page
- FCC — Human Exposure to Radio Frequency Fields: Guidelines for Cellular Antenna Sites
- FCC — RF Safety FAQ
- ICNIRP — Guidelines for Limiting Exposure to Electromagnetic Fields (100 kHz to 300 GHz), 2020
- Health Canada — Safety Code 6
- World Health Organization — Electromagnetic Fields and Public Health: Mobile Phones
- National Cancer Institute — Cell Phones and Cancer Risk Fact Sheet
- American Cancer Society — Do Cell Phones Cause Cancer?
- Germany, Federal Office for Radiation Protection (BfS) — Statement on the Ramazzini Institute Study
- U.S. National Toxicology Program — Cell Phone Radio Frequency Radiation Studies
- FTC — Settlement: Sellers of Bogus Cell Phone Radiation Protection Patches
- FTC — Charges: Cell Phone Radiation Protection Patch Sellers Making False Claims
- Safe Living Technologies — Safe & Sound Pro II product page (manufacturer-published ±6dB accuracy spec)
- ARPANSA (Australia) — Wireless Technologies and Radiation Protection
- BioInitiative Report — Conclusions (source of the precautionary cautionary-target figures)
- BioInitiative Report — background, including the Health Council of the Netherlands' 2008 review
- Building Biology Institute — SBM Evaluation Guidelines (Salzburg-derived precautionary figures, stated in µW/m²)
- Congressional Research Service — Legal Challenges to the FCC's 5G Small Cell Siting Order
- Cellular Phone Taskforce v. FCC, 217 F.3d 72 (2d Cir. 2000)
- Davis Wright Tremaine — Summary of City of Portland v. FCC, 9th Circuit (2020)
- FCC — Fact Sheet on Section 704 of the Telecommunications Act of 1996
This page summarizes and links to primary sources; it does not replace them. Where a study or agency position is contested among researchers, that disagreement is described rather than flattened into a false consensus in either direction.
08 · FAQ
Common questions
"But I feel symptoms near WiFi routers / cell towers — doesn't that prove something?"
Self-reported "electromagnetic hypersensitivity" symptoms are real to the people experiencing them, but controlled double-blind provocation studies — where participants can't see or know when a real RF source is active versus a sham/inactive one — have consistently found people cannot detect the presence of RF fields at levels found in daily life better than chance. That doesn't mean the symptoms aren't real; it means the RF field itself has not been shown to be the cause in controlled testing conditions.
"5G is a completely new, untested technology — shouldn't that change the calculus?"
5G uses a mix of frequencies, some overlapping with previous cellular generations and some new higher (millimeter-wave) bands. The ICNIRP 2020 guidelines were specifically updated to add new restrictions for frequencies above 6 GHz to account for 5G's use of these bands, including tighter averaging areas — this was done deliberately in advance of wide 5G rollout, not skipped.
"Why do some doctors and scientists disagree with the mainstream safety position?"
They do — science isn't unanimous on almost anything, and a genuine minority of researchers argue current limits should be more precautionary. That's a legitimate part of ongoing scientific debate. What's covered on this page is the position of the major, independent regulatory and public health bodies (FCC, ICNIRP, WHO, NCI, ACS, Health Canada, ARPANSA) who have each reviewed the full evidence base and reached similar conclusions — not a claim that zero dissent exists anywhere in the scientific community.
"If consumer RF meters are unreliable, how should I evaluate RF levels at all?"
For almost anyone, the honest answer is: you don't need to. The regulatory limits already contain a large safety margin below any established effect, and ambient RF levels in ordinary environments measure hundreds to thousands of times below those limits per FCC field data. If a specific, credentialed engineering concern exists (e.g. occupational RF exposure near broadcast transmitters), it should be assessed with calibrated, frequency-selective professional equipment and a qualified RF engineer — not a $200–$600 handheld consumer device.
"A city council presentation cited a '10 microwatt' safe limit — was that legitimate?"
Ask which "10 microwatt" figure and in which unit. If it's presented as µW/cm² and attributed to a body like the FCC or ICNIRP, that's not accurate — no mainstream regulator uses that figure. If it's the Salzburg-derived 10 µW/m² precautionary target (a completely different, much smaller number once converted to the same units as the FCC limit), it's a real figure, but from a precautionary advocacy guideline, not a demonstrated harm threshold — see the "10 Microwatt" Claim section above.
"Doesn't federal law mean my city has no say over a 5G tower near me?"
No — your city retains zoning, placement, height, aesthetic, and safety-code authority over the installation. What federal law removes is the ability to deny the application specifically because of RF-emissions health concerns, once the carrier shows FCC compliance. Those are different things, and conflating them is where most of the "cities have zero power" claims go wrong. See the Local Government Authority section above for the actual statute and case law.