Do MRIs Cause Cancer? The Evidence on MRI Safety

MRI scanners do not use ionizing radiation, the type of energy linked to cancer from X-rays and CT scans, and no epidemiological study has established that undergoing MRIs increases your lifetime cancer risk. The magnetic fields and radio waves involved are fundamentally different from the radiation that damages DNA in the way known to seed tumors. That said, the question has more layers than a simple “no,” because recent lab work has found that MRI-strength fields can cause measurable biological changes at the cellular level, and the contrast dyes sometimes injected during scans carry their own set of concerns.

How MRI Energy Differs From Cancer-Causing Radiation

The core reason MRI is considered safe comes down to the type of energy it uses. Ionizing radiation, the kind produced by X-ray machines, CT scanners, and radioactive materials, has enough energy to knock electrons off atoms. When that happens inside your cells, it can produce clusters of damage across one or two turns of the DNA double helix. These clustered breaks are difficult for your cells to repair cleanly, and botched repairs are a well-understood pathway to mutations and, eventually, cancer.

MRI uses three things: a powerful static magnetic field, rapidly switching gradient fields, and radiofrequency (RF) pulses. None of these carries enough energy to strip electrons from atoms. Any DNA lesions that do occur from MRI exposure are far more likely to be isolated, simple breaks rather than the complex clustered damage ionizing radiation causes, and cells repair those simple breaks with high fidelity.¹ This distinction between the quality of DNA damage (not just the quantity) is central to understanding why decades of MRI use have not produced detectable cancer signals in the population.

What Epidemiological Studies Have Found

When researchers have looked at populations of people exposed to strong static magnetic fields, including MRI workers and patients who undergo frequent scans, no clear increase in cancer rates has emerged. A review of the epidemiological evidence on static magnetic field exposure found no indication of increased cancer risk, though the authors were careful to note that most studies involved small numbers and crude methods for estimating how much field exposure people actually received.² The honest summary is that the available evidence does not suggest a cancer link, but the studies conducted so far are not large or precise enough to rule out very small risks with certainty.

This is a common situation in safety science: absence of evidence is not identical to evidence of absence. The practical takeaway, though, is that if MRI caused cancer at any rate even approaching the risk from repeated CT scans, it would almost certainly have shown up by now in the tens of millions of scans performed annually worldwide.

DNA Damage in the Lab vs. DNA Damage That Matters

Some of the more attention-grabbing headlines about MRI safety come from laboratory studies that detect DNA changes after exposure to MRI-strength fields. In one study of human subjects undergoing cardiac MRI, researchers measured a biomarker of DNA double-strand breaks (small protein clusters called γH2AX foci that form at break sites) and found a roughly threefold increase in these markers immediately after the scan.³ That sounds alarming in isolation, but two things put it in perspective.

First, the same study compared the MRI results with DNA damage markers after a chest CT scan and found that the magnitudes were in a broadly similar range: both produced small increases in the same biomarker. The authors characterized both risks as very low. Second, the type of DNA damage matters enormously. As described earlier, MRI-associated lesions tend to be simple, isolated breaks that cellular repair machinery handles routinely. Your cells fix thousands of such breaks every day from normal metabolism alone.

Separately, in vitro work has shown that static magnetic fields at MRI-relevant strengths can boost expression of heat shock proteins, a family of stress-response molecules cells produce when they sense environmental disruption. One study reported up to a 3.5-fold increase in heat shock gene expression after 48 hours of continuous exposure.⁴ That is a genuine biological effect, but heat shock proteins are part of the cell’s normal protective toolkit, not evidence of damage spiraling toward cancer. The finding confirms that MRI fields are not biologically inert, while falling well short of indicating a cancer mechanism.

The Gadolinium Contrast Question

Not every MRI involves contrast dye, but when your doctor needs sharper images of blood vessels, tumors, or inflammation, you may receive an injection of a gadolinium-based contrast agent (GBCA). These are the area where MRI safety concerns have the most substance, and the evidence is genuinely mixed.

A systematic review that pooled 17 studies on gadolinium’s effects on mammalian cells found that 11 of them reported positive genotoxic effects, meaning gadolinium appeared to induce DNA damage in cell cultures. Most of those studies were rated moderate or strong in quality. The reviewers concluded that gadolinium exposure may induce genotoxicity, particularly at high doses or with prolonged exposure, and recommended that MRI scans using gadolinium contrast be prescribed only when clinically necessary.⁵

Before that triggers too much alarm, another study that specifically looked at DNA damage and cell survival when gadolinium was combined with radiation found a more reassuring picture. While initial DNA damage increased, about 90% of that damage was repaired within 24 hours, and there was no statistically significant decrease in the cells’ ability to survive and form colonies.⁶ That repair finding matters because cancer requires not just DNA damage but damage that persists and accumulates in ways the cell cannot correct.

A separate concern is that gadolinium does not always leave the body completely. Deposits have been measured in the brain and bones of patients with normal kidney function, a phenomenon researchers have termed “gadolinium storage condition.”⁷ Whether these deposits cause any long-term harm, including cancer, remains unknown. The deposits were discovered relatively recently, and long-term follow-up data are still accumulating. For people with severe kidney disease, gadolinium carries an additional, well-documented risk: a rare but serious condition called nephrogenic systemic fibrosis, which involves thickening and scarring of the skin and connective tissue. Newer macrocyclic contrast agents have largely mitigated this risk by being more stable and cleared more efficiently, though the bulk of the safety evidence in kidney patients comes specifically from one linear ionic agent that has a partial backup clearance route through the liver.⁸

MRI During Pregnancy

Pregnant women sometimes need imaging for conditions where ultrasound is not sufficient, and MRI is generally preferred over CT because it avoids ionizing radiation. The evidence supports this preference, though with some caveats. A comprehensive review of fetal MRI safety found no conclusive evidence of adverse effects on fetal development at the standard field strengths of 1.5 or 3 Tesla, though caution is still advised in the first trimester when the fetus is most vulnerable to developmental disruption.⁹

A large population-based study from Canada compared outcomes in over 1,700 pregnancies where MRI was performed in the first trimester with more than 1.4 million unexposed pregnancies. The risk of stillbirth, congenital anomalies, cancer, or vision and hearing loss was not significantly higher in the MRI group.¹⁰ The story with gadolinium during pregnancy is less settled. Some early data suggested possible associations with stillbirth and rheumatological conditions in the child, but later research and animal studies have not confirmed those links. Most guidelines still recommend avoiding gadolinium during pregnancy unless the diagnostic benefit clearly outweighs the theoretical risk.

The Risks That Are Real

While cancer is not a demonstrated risk of MRI, the procedure does have genuine safety concerns that are well-documented. These tend to be acute and mechanical rather than long-term and biological.

Noise-Induced Hearing Damage

MRI scanners are loud. The rapid switching of gradient coils produces intense banging and buzzing that can exceed 100 decibels, comparable to a power tool or a rock concert at close range. Without hearing protection, repeated or prolonged exposure could contribute to noise-induced hearing loss. Standard earplugs reduce the perceived sound by roughly 25 to 28 decibels in the frequency range where MRI noise is most intense, and combining earplugs with earmuffs improves that to about 39 to 41 decibels. Researchers have shown that adding a specialized acoustic helmet on top of both can push attenuation to 55 to 63 decibels in the critical range.¹¹ For routine clinical scans, the earplugs or headphones you are given at the imaging center are generally adequate, but researchers who undergo frequent functional MRI sessions for studies should pay extra attention to their cumulative noise exposure.

Metallic Implants and Foreign Bodies

The strong magnetic field means any ferromagnetic metal in or on your body is a potential hazard. Loose metal objects in the room can become high-speed projectiles, and metallic implants inside the body can heat up, move, or malfunction. Radiofrequency energy from the scanner’s transmit coil and the switching gradient fields induce electrical currents in any conductive object inside the bore, and those currents can heat surrounding tissue to dangerous levels.¹² This is why implants remain a contraindication for MRI unless they have been specifically tested and labeled as MRI-compatible, and why the screening questionnaire you fill out before every scan asks detailed questions about surgical hardware, shrapnel, and other metal in your body. Even bulk metallic implants like joint replacements, which are not ferromagnetic and will not fly across the room, can heat up through gradient-induced currents.¹³

Tattoo Burns

Tattoo pigments sometimes contain ferromagnetic metallic compounds, particularly iron oxide. During an MRI, these particles can interact with the electromagnetic field and generate enough local heat to cause a skin burn. A case report involving a professional football player documented exactly this kind of reaction, with the burn attributed to the iron oxide in the tattoo pigment creating small electric currents that raised the local skin temperature.¹⁴ These incidents are uncommon, and most people with tattoos undergo MRI without any problem, but it is worth mentioning to your technologist if you have large or heavily pigmented tattoos.

Ultra-High-Field Scanners and Emerging Concerns

Most clinical MRI scanners operate at 1.5 or 3 Tesla. Research institutions and some specialized clinical centers now use 7-Tesla machines, which produce dramatically sharper images but come with enhanced versions of the biological effects seen at lower field strengths. At 7T, transient bioeffects are more pronounced, and the risk of heating incidents around conductive implants is increased.¹⁵

A more immediately noticeable issue with ultra-high-field scanners is vertigo. Patients and technologists commonly experience dizziness, nystagmus (involuntary eye movement), and imbalance when entering a 7T magnet. These symptoms are caused by the magnetic field’s effect on the vestibular system in your inner ear. Research has shown that simply slowing down the entry into the scanner, taking about two minutes instead of a quick slide in, significantly reduces or eliminates vertigo symptoms.¹⁶ These transient effects resolve once you leave the scanner and have no known long-term consequences, but they are a reminder that stronger fields produce stronger biological interactions, even if those interactions remain non-ionizing.

Why Standard Safety Metrics May Not Tell the Whole Story

The primary measure used to regulate RF energy exposure during MRI is the Specific Absorption Rate, or SAR, which essentially quantifies how much energy your tissues absorb as heat. Regulators set SAR limits to prevent tissue from heating to harmful levels. But some researchers have argued that SAR is an incomplete safety metric. It was designed to capture thermal effects, while most of the recorded biological effects from non-ionizing electromagnetic fields appear to be non-thermal. Even if SAR could be accurately estimated for an entire organ or body, the biological effect may be determined by small amounts of energy absorbed at the level of specific molecules, which current methods cannot calculate.¹⁷

This does not mean MRI is secretly dangerous. It means the safety framework we use to regulate it was built around the most obvious mechanism of harm (heating tissue) and may not fully capture subtler biological interactions. Whether those subtler interactions matter for human health at the doses and durations typical of clinical MRI remains an open question. The cellular stress responses and transient DNA markers discussed earlier are likely part of this picture, but translating lab-dish findings to real-world patient risk is a leap that has not been justified by population data.

The Underappreciated Risk of Incidental Findings

If you go looking for something unrelated to cancer risk from MRI, the most consequential “harm” from the procedure may be what it finds rather than what it does to your body. MRI produces exquisitely detailed images, and those images frequently reveal abnormalities the scan was not ordered to find. In population-based research, whole-body MRI scans turn up incidental findings in a large fraction of participants. These findings often lead to additional testing: more imaging, blood work, biopsies. A study tracking a general population cohort found that disclosure of incidental findings from MRI was associated with a higher number of subsequent biopsies, with factors like older age, female sex, and prior cancer history amplifying the effect.¹⁸

Many of those downstream investigations end up revealing nothing serious. A UK study examining the cascade of care following musculoskeletal MRI found that about two-thirds of patients who were referred for specialist procedural opinions after their scan were ultimately not offered any intervention and were advised to manage their conditions conservatively.¹⁹ This cascade of unnecessary referrals, biopsies, and anxiety is not unique to MRI; it happens with any sensitive imaging modality. But the sheer resolution of modern MRI means it is particularly good at finding things that look worrying on a screen but turn out to be harmless. For patients who undergo frequent surveillance MRI, the psychological burden and procedural risks of chasing incidental findings may represent a more tangible harm than any theoretical carcinogenic effect of the scan itself.

Peripheral Nerve Stimulation

One effect nearly every MRI patient has experienced, even if they did not know what caused it, is peripheral nerve stimulation from the rapidly switching gradient fields. The changing magnetic gradients induce small electric fields in your body, and if those fields are strong enough, they can trigger nerve impulses, producing a tapping, twitching, or tingling sensation, typically in the arms or legs. Researchers have built computational models that predict where and when these nerve stimulations will occur, and the experimental thresholds match model predictions closely, generally within about 5% for most coil configurations and waveforms.²⁰ Scanner manufacturers set gradient slew rates to stay below these thresholds, but some patients, particularly those scanned with high-performance research sequences, still feel mild tingling. The sensation is harmless and stops immediately when the gradient sequence ends. It has no relationship to cancer or lasting tissue damage, but it can be startling if you are not expecting it.