5G and health: what the radiofrequency evidence actually shows

What 5G actually is

5G is the fifth generation of mobile-telecommunications standards. It uses radiofrequency electromagnetic fields — non-ionising radiation, the same broad category as 4G, 3G, Wi-Fi, AM/FM radio, and microwaves — at frequencies ranging from around 600 MHz at the low end to about 6 GHz in the mid-band, and 24–40 GHz in the millimetre-wave bands deployed in some locations.

It is genuinely different from 4G in two ways that matter for a clinical conversation: it transmits at lower individual-device power on average (efficient antenna design plus beamforming targeting specific devices rather than broadcasting in all directions), and it uses higher-frequency bands than older generations. Millimetre-wave frequencies (above 24 GHz) penetrate the body less deeply than the 700 MHz–2 GHz range used by older mobile networks — energy is deposited mostly in the outer skin layers rather than reaching deeper tissue.

The Australian regulator for this — ARPANSA, the Australian Radiation Protection and Nuclear Safety Agency — sets exposure limits that follow the ICNIRP 2020 guidelines. Those limits cover all RF-EMF, including 5G, and are designed with a safety margin of about 50× below any level shown to cause tissue heating.

A. Core clinical — what RF-EMF can and cannot do at relevant exposures

The mechanism of biological effect from RF-EMF at the exposure levels people encounter from mobile networks is well-characterised: gentle warming of tissue. The regulatory limits exist to keep that warming below the threshold for any tissue damage.

ICNIRP 2020 sets these limits with two operative quantities — specific absorption rate (SAR) for whole-body and localised exposures, and absorbed power density for high-frequency 5G millimetre-wave bands. Mobile phones, base stations, Wi-Fi routers, and 5G transmitters operate well below these thresholds in normal use. Compliance is tested and reported.

At those exposure levels:

• No reliable non-thermal mechanism of harm has been demonstrated in the human studies that have attempted to find one. Provocation studies of self-reported electromagnetic hypersensitivity (which is a real lived experience for some people) have not, under blinded conditions, found that subjects can detect when RF-EMF is on or off, nor that their symptoms correlate with actual exposure rather than perceived exposure. The symptoms are real; the proposed cause has not held up.
• Heavy mobile phone use and glioma — the 2011 IARC classification of RF-EMF as Group 2B was largely driven by the Interphone study, which found a hypothesis-generating signal in the highest 10% of mobile-phone-use exposure but inconsistent results across study populations. A more recent 2024 systematic review by Karipidis et al., commissioned by the WHO and conducted at ARPANSA, did not find an association between mobile phone use and glioma at the population level.
• The 2018 NTP rat study is sometimes cited as evidence of harm. The relevant findings were small numbers of malignant heart-tissue tumours (schwannomas) in male rats exposed whole-body to RF-EMF at levels approximately 50× higher than the AU regulatory limit for two years. Female rats and mice showed no consistent effect. The exposure level used in the study is not encountered in real-world human use; the methodology is debated; the FDA, NIEHS, ARPANSA, and SCENIHR have concluded the study does not provide evidence of risk at human exposure levels.

B. Evidence appraisal and the IARC Group 2B context

The Group 2B classification is the most-cited and most-misunderstood piece of information in this conversation.

IARC classifications:

• Group 1 — Carcinogenic to humans. Tobacco, asbestos, processed meat, ionising radiation.
• Group 2A — Probably carcinogenic to humans. Red meat, shift work involving circadian disruption, glyphosate.
• Group 2B — Possibly carcinogenic to humans. This is where RF-EMF sits. Other Group 2B agents include coffee (later reclassified to Group 3 — not classifiable), pickled vegetables, aloe vera leaf extract, lead, gasoline engine exhaust at low concentrations, and chloroform.
• Group 3 — Not classifiable as to carcinogenicity.
• Group 4 — Probably not carcinogenic. (Only one agent has been placed here.)

A Group 2B classification reflects limited evidence in humans and less-than-sufficient evidence in animals. It is an instruction to do more research, not a finding of established carcinogenicity. The popular framing “5G causes cancer” is not what 2B means; nor does it mean RF-EMF is safe — it means the evidence at the time of classification was inconclusive in the direction of low signal.

In 2024, a WHO-commissioned systematic review of 63 studies on mobile phone use and brain tumours found no association — meaningful update to the body of evidence underlying the 2011 IARC decision. IARC may revisit the classification in future monographs; as of 2026 the 2B classification stands but with weakening empirical support.

Other contested specific claims:

• “5G suppresses immune function” — no reliable peer-reviewed mechanistic or clinical evidence supports this at relevant exposure levels.
• “5G activated COVID-19” — biologically implausible (viruses are not activated by radio waves) and not supported by any credible source. ARPANSA, WHO, and major scientific academies have addressed this directly.
• “5G millimetre waves resonate with oxygen and cause hypoxia” — based on a misreading of atmospheric oxygen absorption frequencies. The absorption peak in question is at 60 GHz; the relevant tissue interaction at this frequency is superficial skin heating, not systemic oxygen disruption. SCENIHR and ARPANSA have addressed this.
• “Cell towers near schools are dangerous” — exposure from base stations is typically very low compared with handheld device use. AU practice is to keep tower exposure within the same ICNIRP framework regardless of proximity to schools, and to consult on siting.

C. Australian operations and regulatory context

In Australia, RF-EMF safety is governed by ARPANSA’s Radio Frequency Standard RPS S-1, which adopts the ICNIRP 2020 guidelines. All Australian carriers and equipment must comply.

(MBS / PBS items verified 2026-05-16 via WebSearch — workspace egress to mbsonline.gov.au + pbs.gov.au still blocked; not directly relevant here as this is environmental rather than medical-services policy.)

Other regulators with adjacent roles:

• ACMA (Australian Communications and Media Authority) — regulates radiocommunications licences and enforces compliance with ARPANSA exposure limits.
• TGA — regulates medical devices that emit RF (e.g. diathermy, MRI, RF-based ablation devices), separately from the telecommunications framework.
• Department of Health — issues public-facing communications about environmental exposures and health.
• NHMRC — funds research and reviews the evidence base periodically.
• Cancer Council Australia — provides patient-facing summaries that align with the regulatory position.

For a patient who is genuinely concerned and asking what to do, the AU-aligned conversation usually covers:

• Source of concern. Specific symptom that the patient attributes to RF? General environmental concern? Family member with a different view? Each warrants a different response.
• Symptom history. If specific symptoms are present (headaches, sleep disturbance, fatigue, palpitations), a clinical workup for the more probable causes — sleep, mood, blood pressure, thyroid, medication side effects — comes first. RF-EMF is not in the top 20 differentials for any of these by current AU evidence.
• Practical reduction. For general thermal-comfort or peace-of-mind reasons, using speakerphone or wired earphones in long calls, keeping phones off the body during the day, and minimising bedroom-side wireless devices are reasonable. None are “necessary” by ARPANSA’s current position; all are zero-harm and may meaningfully reduce a person’s worry.
• What to avoid. Spending substantial money on unverified shielding products, “EMF protection” jewellery, or unproven supplements does not reflect a primary-tier conclusion about RF-EMF risk and can divert resources from more impactful changes.

D. Where the honest uncertainty sits

The 2026 picture, as I read it:

1. The exposure science is mature. ICNIRP 2020 reflects 30 years of cumulative work. The thermal mechanism is well-characterised. The current AU regulatory limits include a substantial safety margin.
2. The epidemiology has weakened the 2011 IARC 2B signal, not strengthened it. The 2024 systematic review is methodologically more robust than the studies that informed the 2011 classification.
3. Specific subpopulations remain under-studied — long-term effects of millimetre-wave-specific exposure, paediatric exposure from cumulative lifetime devices, exposure during pregnancy. ARPANSA and ICNIRP both note these as ongoing research priorities.
4. The marketing narratives — on both sides — have run ahead of the evidence. Confident claims of harm and confident claims of “completely safe” are both more certain than the science actually justifies. The middle position — “no established adverse effect at regulated exposure levels; some open research questions; precautionary measures available for those who want them” — is the position the AU and international scientific communities hold.

For a patient or family asking “should I be worried about 5G?”: the honest answer is no, at current Australian exposure levels, by the standards of current evidence. That answer can coexist with reasonable individual preferences for lower-exposure practices.

What this article is and is not

This is general health information drawn from current Australian regulatory positions (ARPANSA, NHMRC, Department of Health, Cancer Council Australia, TGA), international guidance (ICNIRP, WHO, IARC), and recent peer-reviewed reviews. It is not personal medical advice and does not create a doctor–patient relationship.

For Australian consumer-friendly sources: ARPANSA — EMF and your health, Cancer Council Australia, and the Department of Health — Mobile communications and your health.

---

Sources cited

1. ARPANSA — Radiofrequency exposure and 5G
2. ARPANSA — Radio Frequency Standard RPS S-1
3. NHMRC — Radiofrequency reviews
4. ICNIRP Guidelines 2020
5. WHO — Electromagnetic fields and public health
6. IARC — Monograph Volume 102
7. TGA — Medical devices
8. Department of Health — Mobile communications and your health
9. Cancer Council Australia
10. NIEHS — Cell phones
11. Karipidis et al. — Mobile phone use and brain tumours systematic review (Environ Int 2024)
12. SCENIHR — EMF health effects
13. Australian Mobile Telecommunications Association