EMF exposure limits are set by scientific bodies that review the research, find the lowest exposure level with an established harmful effect, and then divide that level by a large safety factor before writing the rule. For the general public, that factor is 50. In other words, the official limits sit fifty times below the established heating threshold. What that recipe does not cover is the research into effects below heating, which is where the entire modern argument lives. Understanding how that process works, and what it is built on, explains most of the arguments people have about EMF.
This is a neutral tour of the machinery. No side-taking, just how the sausage gets made.
Key Takeaways
- Limits start from the lowest exposure with an established adverse effect, then divide by a safety factor: 10 for workers, 50 for the public.
- For RF, the established effect the limits are built on is heating of tissue. That is called the thermal basis.
- ICNIRP writes the guidelines most of the world uses. The FCC sets the legally binding US rules. They arrive at similar numbers.
- Research into possible effects below the heating level continues. That open question is where the whole precaution debate lives, on both sides.
Who Actually Sets the Limits
Two names come up constantly on this site, and they do different jobs.
ICNIRP, the International Commission on Non-Ionizing Radiation Protection, is an independent scientific commission recognized by the World Health Organization. It reviews the research and publishes exposure guidelines. Most countries adopt ICNIRP’s numbers, or something close to them, into their own law. Its current RF guidelines were updated in 2020, partly to cover the frequencies 5G uses.
The FCC, the Federal Communications Commission, sets the legally binding limits in the US. The FCC is not a health agency; it adopted its current limits in 1996, based on recommendations from scientific and health bodies of that era, including IEEE and the EPA-era radiation councils, with input from the FDA on phones. Its 2019 review reaffirmed those limits without a new derivation, a fact that matters to both sides of this argument and belongs in any honest account of it. ICNIRP’s history runs differently: guidelines in 1998, and a 2020 replacement of the radio-frequency portion covering 100 kHz to 300 GHz, with the substantive changes above 6 GHz to address newer bands. The FCC and ICNIRP numbers differ slightly in structure but land in the same range.
The Recipe, Step by Step
Strip the process to its skeleton and it has four steps.
- Review the research. Thousands of studies get evaluated for quality and consistency. One alarming result does not move a limit; a reproducible, established effect does.
- Find the threshold. Identify the lowest exposure level at which an established adverse effect appears. For RF, the anchor is a whole-body heating effect observed around 4 W/kg, a measure of absorbed energy per kilogram of body tissue.
- Divide by the safety factor. Workers, who are trained and monitored, get a factor of 10. The general public, which includes children, the elderly, and the sick, gets a factor of 50. That is how 4 W/kg becomes a public whole-body limit of 0.08 W/kg.
- Translate into measurable numbers. Nobody can measure watts per kilogram of their own body at home, so the bodies convert that absorption limit into field strengths and power densities a meter can read, called reference levels. Those are the numbers I compare readings against in What Is a Safe EMF Level?.
The unit in step two has a name you have seen on phone boxes: SAR, specific absorption rate. Phone testing uses a localized version of the same idea. In the US, a phone must stay under 1.6 W/kg in the most exposed gram of tissue; ICNIRP’s version allows 2 W/kg averaged over ten grams. Different math, same recipe.
The Thermal Basis, and Why It Matters
Here is the single most important fact for understanding EMF arguments: the established effect these RF limits are built on is heating. Strong RF warms tissue, the way a microwave warms food, and the science on when that happens is solid. The limits are engineered to keep exposures far below the point where meaningful heating begins.
What about effects that do not involve heating? That is the active research frontier. Studies on possible non-thermal effects continue, and the results are why this debate is real rather than manufactured. The US National Toxicology Program’s large animal study reported clear evidence of tumors in male rats at high exposures; the Ramazzini Institute reported findings in the same direction; the WHO’s cancer agency, IARC, classified RF as Group 2B, “possibly carcinogenic,” in 2011. So far, the bodies that set limits have judged this evidence insufficient to move them, ICNIRP’s 2020 review said exactly that, and researchers who disagree have said so in the literature. That is an accurate picture of a live scientific argument, not a settled one.
Both sides of the EMF argument grow out of this one spot. One side says: the limits cover everything that is established, with a 50x margin, so relax. The other says: the limits are keyed to heating, the non-thermal questions are still open, so leave extra room. Notice that neither of those positions requires the other to be lying. My own position, and the reasoning behind it, lives in The Precautionary Principle, Explained Fairly.
Honest Footnotes
A fair tour includes the disputes, so here they are, stated plainly.
The FCC reviewed its limits in 2019 and kept them. In 2021, a US appeals court ruled the FCC had not adequately explained parts of that decision, particularly regarding evidence on non-thermal effects, and sent it back for better reasoning. The court did not rule the limits unsafe, and it did rule that the agency failed to give a reasoned explanation for keeping them, including with respect to non-thermal evidence in its own docket. Both halves of that sentence matter, and readers deserve both.
Some countries, and some cities like those in Switzerland and Italy, apply stricter installation limits than ICNIRP recommends, usually as policy caution rather than a different reading of the health science. I compare those in EMF Limits by Country.
And ICNIRP has critics who argue its membership and thermal focus are too narrow, just as the critics have critics who note ICNIRP’s process is the most reviewed in the field. I flag that the argument exists; resolving it is above this site’s pay grade, which is precisely why I measure instead.
What This Means When You’re Holding an EMF Meter
The practical takeaway is context. One foot from my own router under load, my meter ran off the top of its scale, which means the level was above 20,000 µW/m². You now know what that number is being compared against: a public limit derived from a heating threshold, divided by 50, of roughly 10,000,000 µW/m². So that reading sits at around 0.2% of the limit.
You also know what the limit is not: a verdict on the open research questions. The same reading sits well above the strictest precautionary targets, covered in Building Biology Standards (SBM-2015), Explained, and both of those statements describe the same number.That is the honest frame this site runs on. The limits tell you what is established. Your meter tells you what is actually in your home. What you do between those two facts is yours.
Frequently Asked Questions
What are the FCC limits for phone radiation?
US phones must test under a SAR of 1.6 W/kg in the most exposed gram of tissue. Every phone sold legally in the US passed that test, and you can look up any model’s results in the FCC’s equipment database. I explain how to read those numbers in Phone SAR Ratings, Explained.
Do the limits cover 5G?
Yes. ICNIRP’s 2020 update explicitly extended and refined its guidelines for the frequency ranges 5G uses, including the higher millimeter-wave bands. Whether existing measurements capture 5G well is a separate, practical question I cover in What 5G Actually Is.
Are US limits stricter or looser than other countries?
For phones, the US localized SAR limit of 1.6 W/kg is nominally stricter than ICNIRP’s 2 W/kg, though the averaging methods differ. For environmental exposure, several countries and regions apply stricter limits than ICNIRP as a policy choice. The full comparison is in EMF Limits by Country.
Who checks that devices actually comply?
Manufacturers must test devices against the limits before sale, and regulators like the FCC certify and can audit the results. Home meters cannot verify SAR, which requires lab equipment, but they can verify the environmental reference levels, which is what I do in every article here.
Now that you know how the official benchmark gets made, the other benchmark deserves the same fair treatment: Building Biology Standards (SBM-2015), Explained covers the strict precautionary targets and what they do and do not claim. The rest of this section lives on the Standards & Regulation hub.
Related Reading
- What Is a Normal EMF Reading in a House?
- What Is EMF? Ionizing vs. Non-Ionizing, in Plain Words
- Does EMF Get Weaker With Distance?
- How to Read an EMF Meter: What the Numbers and Units Mean