The FCC's 2019 RF decision kept the limits and rewrote the paperwork
In FCC 19-126, adopted November 27, 2019 and released December 4, 2019, the FCC closed the review it had opened in 2013 and left the 1996 exposure limits exactly as they were. What it changed was the machinery around them, starting with which transmitters need a formal RF evaluation and what operators must do near antennas. On the health question it relied heavily on the FDA, and that reliance is the part a federal court later found too thin.
The full order runs 159 pages. Much of it is engineering detail. The parts below are the ones that still decide what applies today.
Four documents in one
The order bundles four actions under one number, with Commissioner Rosenworcel concurring:
- A Resolution of Notice of Inquiry closing ET Docket 13-84, the 2013 review of whether the limits were still adequate.
- A Second Report and Order in ET Docket 03-137, rewriting how compliance is shown.
- A Notice of Proposed Rulemaking opening ET Docket 19-226 on newer technology.
- A Memorandum Opinion and Order denying a petition for reconsideration about outer ears.
Only the second one changed rules that bind anyone right now. The first is the part the D.C. Circuit sent back in 2021.
Why the FCC kept the limits
The inquiry drew more than 1,000 comments and ex parte filings. The FCC wrote that "the vast majority of filings were unscientific" and that even the scientific ones did not make a persuasive case or offer a specific alternative limit backed by rigorous analysis. Some commenters pointed to the BioInitiative Report and Building Biology guidelines, which call for limits the FCC described as "millions to billions times more restrictive" than its own. The FCC's view was that no device could reliably transmit usable signals at those levels.
It also declined to loosen anything. Some commenters asked the FCC to move to the IEEE or ICNIRP standards, which allow 2 W/kg averaged over 10 grams, and the order refused for the same reasons it refused to tighten.
On phone testing, the FCC kept its procedures. Phones are tested against a model head with no gap and, for body-worn use, at up to 2.5 cm from the body. Hotspot mode is tested at 1 cm. The order argued that testing at full power, with what it called "a large safety margin" built into the limits, made zero-spacing tests unnecessary. On children, it pointed to the IEEE 1528-2003 test setup, which it said was built to represent children as well as adults, and to an FDA statement that the evidence did not show danger to children and teenagers.
The order addressed the 2018 National Toxicology Program rat study in a footnote. It quoted an NTP scientist who said the results could not be applied directly to people because the animals were exposed over their whole bodies, at higher levels and for longer than people are, and that the studies used 2G and 3G phone signals.
What actually changed: who has to run an evaluation
Before 2019, whether a transmitter needed a routine RF evaluation depended on its service category. The FCC called the old rules "a patchwork of outdated and inconsistent rules" and gave an example: a Part 101 millimeter-wave transmitter and a Part 30 transmitter with similar power and antenna gain were treated differently, one exempt and one not.
The new rules apply the same three tests to every service:
- A blanket exemption for sources with no more than 1 mW of time-averaged power, at any distance, from 100 kHz to 100 GHz.
- A SAR-based formula for antennas from 0.5 cm to 40 cm from the body, valid from 300 MHz to 6 GHz.
- An MPE-based table of power thresholds for everything else, from 300 kHz to 100 GHz, usable only at distances of at least λ/2π from the antenna.
The MPE table is simple enough to run yourself. Between 300 and 1,500 MHz the threshold ERP in watts is 0.0128 × R² × f, with R in meters and f in MHz. From 1,500 MHz to 100 GHz it is 19.2 × R². Here is what that gives for a transmitter 10 meters from the nearest person.
Assumes R = 10 m and time-averaged ERP. 100 MHz uses the 30 to 300 MHz formula, 3.83 × R². The FCC's own estimate was about 1 kW at 850 MHz and 2 kW at 1900 MHz.
| Frequency | Threshold ERP |
|---|---|
| 100 MHz | 383 W |
| 600 MHz | 768 W |
| 850 MHz | 1,088 W |
| 1900 MHz | 1,920 W |
| 3.7 GHz | 1,920 W |
| 28 GHz | 1,920 W |
Source: Our arithmetic from the MPE-based exemption table adopted in FCC 19-126, now in 47 CFR 1.1307(b)(3).
Our results line up with the FCC's own check: it said a base station at roughly ten meters could run up to about 1 kW ERP at 850 MHz or 2 kW at 1900 MHz and stay exempt, close to what the old table allowed. The FCC expected most sites that were exempt before to remain exempt.
Exemption never meant immunity. The order repeats several times that every licensee stays responsible for meeting the limits whether or not it had to run the evaluation.
Mitigation, and when it all took effect
The Second Report and Order also spelled out what operators must do where people could get close to an antenna: signs, access control, training for workers, and supervision of untrained people allowed into restricted areas. Licensees got two years from the effective date to re-check sites that had been exempt under the old rules.
That effective date slipped. The rules ran in the Federal Register on April 1, 2020, at 85 FR 18131. On June 2, 2020 the FCC delayed the key sections indefinitely because they needed Office of Management and Budget approval under the Paperwork Reduction Act. The amended sections 1.1307, 2.1091 and 2.1093 finally took effect on May 3, 2021, and the FCC confirmed the two-year clock started that day. By our count the window closed in May 2023.
What changed and what stayed the same
| Item | Kept or changed | Detail |
|---|---|---|
| Public SAR limit | Kept | 1.6 W/kg over 1 g; 0.08 W/kg whole body |
| Extremities, including outer ears | Kept | 4 W/kg over 10 g; petition to change the ear treatment denied |
| MPE limits in the air | Kept | Same Table 1 values, 300 kHz to 100 GHz |
| Phone test positions | Kept | Head with no gap; body up to 2.5 cm; hotspot 1 cm |
| Policy on children | Kept | FCC declined to revisit |
| Who must evaluate | Changed | Three formula tests replace service-by-service rules |
| Very low power sources | Changed | Blanket 1 mW exemption at any distance |
| Signs, access, training | Changed | Specific mitigation duties, two-year transition |
| Localized limit above 6 GHz | Proposed only | 4 mW/cm² over 1 cm², up to 3 THz |
| Wireless power transfer | Proposed only | Exposure limits for wireless charging equipment |
On the last two rows: the NPRM in Docket 19-226 also proposed limits on electric fields induced inside the body at low frequencies, modeled on ICNIRP's 2010 guidelines. The current text of section 1.1310, last amended in April 2020, still has no localized power density limit, and its table still stops at 100 GHz. We did not find an order adopting the other proposals either.
The FDA's role
The FCC says plainly that it is not a health and safety agency. In 1996 it built its limits on recommendations from the EPA, the FDA and other federal agencies. The FDA runs the federal radiation control program for electronic products under 21 U.S.C. 360ii, and on its cell phone pages it says it "shares regulatory responsibilities for cell phones" with the FCC, which sets the emission limits.
Before closing the inquiry, the FCC asked the FDA on March 22, 2019 whether the science called for changes. The FDA's April 24, 2019 reply said "no changes to the current standards are warranted at this time." The order quoted that letter, an FDA web page and a 2018 FDA statement. In 2021 the D.C. Circuit called all three conclusory, because they stated a result without walking through the studies.
Two months after the order, in February 2020, the FDA published a detailed review of its own, limited to cancer.
- Feb 2020
- FDA literature review published, two months after FCC 19-126
- About 125
- Animal studies weighed, published 2008 to August 2018
- About 70
- Epidemiological studies weighed, published 2008 to May 2018
- Tumors only
- The review's stated scope
The review concludes that "there is insufficient evidence to support a causal association between RFR exposure and tumorigenesis." It is open about its limits. Exposure in the human studies is mostly self-reported, there are no direct measurements outside the lab, and one section is titled "Lack of Evidence Based on the U.S. Population." It notes that some studies suggest a link in heavy users but finds no clear and consistent pattern.
Its scope matters for the remand. The FDA says the analysis "was scoped to assess any possible causal relationship between RFR exposure and the formation of tumors." The court upheld the FCC on cancer. The questions it sent back are about effects other than cancer, plus children, long-term exposure, testing and the environment. A tumor review, however careful, does not speak to most of those on its face. The FCC's September 2026 public notice on the remand cites the review in a footnote.
What the order said about reducing exposure
One footnote is worth knowing. The FCC wrote that it "does not endorse the need to take measures to further reduce exposure," but that for consumers skeptical of the science it keeps a page of simple steps. Retail products sit outside that guidance. Shops sell items such as EMF stickers for phones, including EMFCALM sticker bundles in packs of 3, 6 and 12; the order never mentions products like these, and we don't evaluate them.
Nearly seven years on, the 2019 order is the most recent FCC order on its limits that we could find. Its compliance rules have been in force since May 2021. The decision behind them, keeping the 1996 numbers, is the one part still waiting for the explanation the court asked for in 2021.